Hong Kong J Psychiatry 2007;17:45-54

ORIGINAL ARTICLE

Dr SK Leung, MBBS (HK), MSc, MRCPsych (UK), FHKCPsych, FHKAM (Psychiatry), Castle Peak Hospital, Tuen Mun, New Territories, Hong Kong, China.
Prof Peter C Spurgeon, BSc, PhD, Health Services Management Centre, University of Birmingham, England, United Kingdom.
Dr HK Cheung, MBBS (HK), FRCPsych (UK), FHKCPsych, FHKAM (Psychiatry), Castle Peak Hospital, Tuen Mun, New Territories, Hong Kong, China.

Address for correspondence: Dr SK Leung, Castle Peak Hospital, 15 Tsing
Chung Koon Road, Tuen Mun, New Territories, Hong Kong, China. Tel: (852) 2456 7111; Fax: (852) 2463 1644;
E-mail: >rskleung@ha.org.hk

Submitted: 27 March 2007; Accepted: 21 May 2007

pdf Full Paper in PDF

Introduction

In recent years, job satisfaction and stress among mental health care workers has become a focus of research, especially in the context of hospital-based and community- based care, as currently there is a shift of focus from the hospital to community.^1-5 Evidently, community-based mental health staff were generally more stressed than their ward-based counterparts, though there was no significant difference in job satisfaction between the two groups.³ Among the health care professions, community-based and ward-based mental health nurses appear to be two of the most stressed groups.⁶

The relationship between stress and job satisfaction among nurses is interesting. A meta-analysis found that job satisfaction of nurses had the strongest negative association with stress.⁷ However, community nursing, including community mental health nursing, appears to be more stressful but also more satisfying than corresponding nursing in hospital settings.^1,8 Job satisfaction of nurses was found to be an important variable associated with turnover of nurses, their productivity and performance.^9-12 The level of nurses’ job satisfaction may influence patient care.^13,14 An understanding of the changing nature of nurses’ job satisfaction may help to solve recruitment and retention problems.¹⁵

Managing motivation and developing job satisfaction is important in the health care environment.¹⁶ Maslow¹⁷ and Herzberg¹⁸ have suggested motivation as a major component in job satisfaction. According to Herzberg’s motivation-hygiene theory, removal of hygiene factors causes job dissatisfaction, while motivators are sources of job satisfaction.¹⁸ Some authors have proposed that intrinsic factors serve as sources of job satisfaction, while extrinsic factors contribute to job dissatisfaction.¹⁹ Based on Maslow’s and Herzberg’s theories, measures of job satisfaction have been developed,^20,21 and studies have been carried out with attempts to identify areas and dimensions of job satisfaction and dissatisfaction among nurses.^12,22,23

Various factors have been associated with job satisfaction of nurses. Age, education, years of experience, clinical grade, communication, recognition, opportunity for advancement, autonomy, commitment, and stress have all been implicated.^9,11,15 Job satisfaction has been studied among nurses in different work settings and specialties.^24,25 A study found that 70% of a sample of community psychiatric nurses (CPNs) rated their satisfaction as ‘high’ or ‘very high’, while only 4% reported a low level of satisfaction.²⁶ Factors suggested as contributing to the job satisfaction of CPNs include: autonomy, patient care (‘seeing clients improve’ and ‘preventing admission’), organisation of service and support of managers, and interaction with members of the primary health care team.²⁶

Another study²⁷ compared nurses’ satisfaction and stress in admission, short-stay and long-stay wards of a psychiatric hospital. They found that although each group of nurses seemed to be generally satisfied with their work, those in the admission ward showed higher levels of job satisfaction, suffered less stress, and enjoyed better health compared to nurses in the other wards.²⁷ It was suggested that the nature of the work was responsible: nurses in the admission ward could get immediate feedback on their efforts and enjoyed regular interaction between themselves and management staff; their nursing skills were perceived as important and further training was encouraged; and difficult- to-manage patients could be transferred to the short-stay ward. On the other hand, nurses in the short-stay ward felt frustrated, dealing with patients difficult to re-socialise and complained about the pressure to accept patients from the admission ward. Nurses in the long-stay ward, however, expected patients’ progress to be slow and had less pressure than those in the short-stay ward. It therefore seemed that nurses in the long-stay ward were more satisfied with their jobs than their counterparts in the short-stay ward, but less satisfied than those in the admission ward.²⁷

Different methods have been employed to study nurses’ job satisfaction. Although content analysis may provide an objective and systematic method for understanding human behaviour,^28,29 it is labour-intensive, time-consuming, and not used extensively.³⁰ More commonly, nurses’ job satisfaction is studied using quantitative instruments usually in the form of self-completed questionnaires. These provide a measure of job satisfaction, which can be correlated with other assessable variables. Some instruments may be generally applicable to all occupations. Examples include the Index of Job Satisfaction,³¹ the Minnesota Satisfaction Questionnaire,²⁰ the Job Satisfaction section of the Job Diagnostic Survey,³² and the Job in General Scale.³³ Some may be specific to the nursing profession, like the McCloskey / Mueller Satisfaction Scale (MMSS),²¹ the Index of Work Satisfaction (IWS),^34,35 and the Measure of Job Satisfaction (MJS).³⁶ Instruments specific to the nursing profession, particularly the MMSS and IWS, are more widely used for studying nurses’ job satisfaction,^{21-23,25,37,38} while the MJS is not commonly used by other researchers as its theoretical basis may be weak.

The IWS consists of 44 statements rated on a 7- point Likert scale and grouped under 6 components, namely, pay, autonomy, professional status, interactions, task requirements, and organisational policies.^34,35 The mean response for all statements within a component is the raw satisfaction score for that component. The IWS is a weighted mean of raw satisfaction scores for the 6 components. The raw scores for each component are weighted by how important the respondents feel that component is to them relative to the 5 other job satisfaction components. The level of importance is assessed through a series of pair-wise comparisons of components, and through certain transformation procedures the component weighting coefficients are obtained.^34,35 The IWS has its strength on the organisational perspective of studying nurses’ job satisfaction, and can be particularly useful for measuring the change in the level of job satisfaction, for instance, before and after implementation of a certain organisational policy.³⁵ It has been used extensively,^25,37,38 and norms obtained from studies are provided for reference and comparison.³⁵ However, the scale requires considerable time and effort for completion, and may deter responses. Moreover, the scoring system is very complicated, which requires reference to a matrix and table or even the use of computer software.^34,35

In an attempt to identify the rewards which keep nurses on the job, McCloskey³⁹ developed a satisfaction scale. This scale was based on the theories of Maslow and Burns, and measured safety rewards (potentially against dangerous threats), social rewards (need to belong), and psychological rewards (autonomy, responsibility, recognition, and appreciation). The measurement characteristics were subsequently rigorously examined by Mueller and McCloskey, who developed the current version of the MMSS.²¹ The MMSS consists of 31 items rated on a 5-point Likert scale and grouped by factor analysis into 8 subscales: satisfaction with extrinsic rewards, scheduling, family / work balance, co-workers, interaction, professional opportunities, praise / recognition, and control / responsibility. A mean score for each subscale and an overall mean score for the global scale can be obtained. The scale was shown to be reliable and valid, and a more valid measure of nursing satisfaction than other scales not designed for nurses.²¹ The MMSS is easy to use, the scoring system is simple compared to the IWS, and has been used in a number of nurses’ job satisfaction studies.^22,23 Most studies on nurses’ job satisfaction were conducted in western countries; only limited local data on this subject are available.

In 1993, a group of researchers from academia and the Hospital Authority in Hong Kong tried to develop a measure of nurses’ job satisfaction suitable for use in the local context.⁴⁰ They refined the MJS and developed a scale with 38 items available bilingually (English and Chinese).^36,40 The principal component analysis yielded 9 factors: communication, professionalism, routinisation, pay and prospect, locus of control, education and training, commitment, personal control, and recognition.⁴⁰ However, this refined scale bears close similarity and shares similar weakness (such as repetition of items) with the MJS, and is not commonly used by other researchers.

Another study on job satisfaction and need for autonomy of registered nurses (RNs) in Hong Kong using the IWS found that their sample was dissatisfied more than satisfied, and valued the components of autonomy, professional status and pay more than interaction, task requirements and organisational policies.^34,41 There was no significant difference in job satisfaction between subjects working in the general and psychiatric units in a chronic hospital.⁴¹ Factors related to stress and coping among nurses in Hong Kong have been studied.⁴² However, studies focusing on job satisfaction among psychiatric nurses, including the relationship of their job satisfaction with other variables (stress and work setting) are relatively lacking. Moreover, areas of job satisfaction and dissatisfaction and whether they can be grouped under certain major factors or dimensions remain unknown.

This study was carried out to describe the job satisfaction among a large cohort of psychiatric nurses in Hong Kong, using standard measuring instruments to determine correlations between job satisfaction and socio- demographic variables, work setting, and stress. It also set out to identify any major factors or dimensions of job satisfaction among psychiatric nurses.

Patients and Methods

Subjects and Setting

This was a cross-sectional study conducted with the psychiatric nurses of Castle Peak Hospital (CPH) in Hong Kong, which serves a population of about 930,000 from Tuen Mun, Yuen Long, and Tin Shui Wai.⁴³ The study was approved by the Ethics Committee of the New Territories West Cluster.

All psychiatric nurses with the rank of: ward manager or service manager (for simplicity, collectively referred to as ward manager), nursing officer (NO), RN, and enrolled nurse (EN) were sent questionnaires on 25 September 2002. The nurses could be ward-based (acute / admission ward or subacute / chronic ward) or community-based and were associated with adult general psychiatric teams serving the 3 geographical catchment areas of CPH. All the recipients were requested to return the completed questionnaires in sealed envelopes marked ‘confidential’ by internal mail before 16 October 2002.

Questionnaires

The questionnaires collected data on the following three main areas:

1. Job satisfaction (MMSS)
   The scale consisted of 31 items grouped under 8 subscales: satisfaction with extrinsic rewards (items 1-3), scheduling (items 4-6, 8-10), family / work balance (items 7, 11, 12), co-workers (items 14, 15), interaction (items 16-19), professional opportunities (items 20, 21, 27, 28), praise / recognition (items 13, 24-26), and control / responsibility (items 22, 23, 29-31).²¹ Each item was scored from 1 to 5 with 1 corresponding to ‘very dissatisfied’ and 5 corresponding to ‘very satisfied’. For each subscale, scores were summed and divided by the number of items to obtain a mean. An overall mean for the global scale could be obtained as a general measure of nursing satisfaction, and has been found to be reliable and valid for measuring nurses’ job satisfaction.²¹
2. Stress (Anxiety Stress Questionnaire)
   The Anxiety Stress Questionnaire (ASQ) consisted of 17 items grouped under 3 subscales: job-induced tension (items 8-10, 12, 14, 16, 17), somatic tension (items 4, 5, 7, 11, 15), and general fatigue and uneasiness (items 1-3, 6, 13).⁴⁴ Responses to each item were true or false, which scored 2 or 1 respectively (except for items 1 and 2 which scored 1 for true and 2 for false), and averaged. A mean score for each subscale and an overall mean score for the global scale were obtained. The internal reliability coefficients for the 3 subscales are 0.83 (job-induced tension), 0.76 (somatic tension), and 0.72 (general fatigue and uneasiness) respectively. The test-retest reliability of the whole scale is 0.79. Concurrent validity of the ASQ has been reported.⁴⁵ The scale has been widely used for nurses and other workers, and has been used in a recent study on stress and coping of Hong Kong nurses.⁴²
3. Background information and work-related variables Socio-demographic data including age, gender, marital status, number of children, and education level, as well as work-related variables, including: years of nursing experience, rank, and work setting (community psychiatric, acute ward, or chronic ward nurse) were collected.

Statistical Analysis

The data were processed and analysed using Statistical Package for the Social Sciences (Windows version 10.0; SPSS Inc, Chicago [IL], US). Descriptive statistics were used for characterising the job satisfaction of psychiatric nurses. For determining the associations and correlations between job satisfaction and other socio-demographic and t tests, one-way analysis of variance, Pearson correlation coefficients) were used, depending on the characteristics of the variables. Significance level was set at p < 0.05. Stepwise multiple linear regression analysis was performed to determine which of the variables might best predict job satisfaction. Maximum likelihood factor analysis with oblique rotation was performed on the items of job satisfaction in an attempt to identify any major factors or meaningful constructs for job satisfaction

Results

Background Data and Stress (Anxiety Stress Questionnaire)

A total of 212 questionnaires were sent to all the ward- based and community-based psychiatric nurses of the general adult psychiatric teams of CPH. One hundred and seventeen questionnaires were returned, giving a response rate of 55.2%. Fifteen of these could not be used for analysis because they were incompletely answered, leaving a final sample of 102 that constituted 48.1% of those sent out. On comparing the 102 respondents with analysable questionnaires and the remainder, there were significantly more ward managers (10 out of 13) and fewer ENs (28 out of 75) in the former group (p < 0.05), but there were no significant differences between the two groups regarding other ranks and work settings. Among the 102 subjects, 43 (42%) were male and 59 (58%) were female. Thirteen subjects did not disclose their age. The mean age of the remaining 89 subjects was 36 (SD, 7; range, 23-53) years. The socio-demographic data and work-related variables of all 102 subjects are summarised in Table 1. The scores of the ASQ are summarised in Table 2.

Job Satisfaction (McCloskey / Mueller Satisfaction Scale)

The scores of job satisfaction are summarised in Tables 3 and 4. If a score of 3 is taken as the neutral point, the results showed that overall the subjects were satisfied with their job (the mean ± SD of the overall mean score of the global scale was 3.24 ± 0.50). They were satisfied in 6 of the 8 subscales. The three subscales revealing most satisfaction were extrinsic rewards (3.74 ± 0.56), co-workers (3.62 ± 0.59), and praise and recognition (3.42 ± 0.68). The nurses were dissatisfied with the balance of family and work (2.86 ± 0.67) and professional opportunities (2.90 ± 0.68). Regarding individual items, they were satisfied with 22 items, dissatisfied with 8 items, and neither satisfied nor dissatisfied with 1 item. The three most satisfying items were salary (4.13 ± 0.68), nursing peers (3.85 ± 0.68), and hours of working (3.74 ± 0.74). The three most dissatisfying items were child-care facilities (2.47 ± 0.93), opportunities for career advancement (2.62 ± 1.02), and opportunity for part-time work (2.76 ± 0.87).

Correlation between Job Satisfaction and Socio- demographic and Work-related Variables and Stress

As some of the subjects did not disclose their age and as age correlated significantly with years of nursing experience (Pearson correlation coefficient [r] = 0.92, p < 0.001) among those who disclosed their age, age was not entered into statistical analysis. Negative correlations were noted between the overall mean score for the MMSS (AMMSSTOT) and the overall mean scores for the global scale of the ASQ (r = –0.31, p = 0.002), and (i) the mean scores for all the subscales of the ASQ, including job- induced tension (r = –0.26, p = 0.01), (ii) somatic tension (r = –0.31, p = 0.002), and (iii) general fatigue and uneasiness (r = –0.24, p = 0.02). The overall mean score of the MMSS was significantly higher in the group of ward managers (mean ± SD, 3.56 ± 0.32) than the group of non-ward managers (i.e. the group of NOs, RNs, and ENs as a whole; mean ± SD, 3.21 ± 0.51) [t = 2.14, df = 100, p = 0.04]. No statistically significant association was found between the overall mean score for the MMSS and gender, marital status, number of children, years of nursing experience, and work setting.

The variables which yielded a significant association with the overall mean score for the MMSS were then subjected to stepwise multiple linear regression analysis. The overall mean score for the MMSS (AMMSSTOT) was best predicted by the mean score for the somatic tension subscale (β = –0.30, p = 0.002) and the rank of ward manager (β = 0.20, p = 0.03). Thus, being a ward manager and having less somatic tension predicted greater overall job satisfaction.

Factor Analysis of Job Satisfaction

The items on the MMSS were subjected to factor analysis to examine if they formed meaningful constructs. Using maximum likelihood factor analysis with oblique rotation (which was consistent with the process used by Mueller and McCloskey),²¹ 3 factors could be identified (Table 5). The eigenvalues of Factors 1, 2, and 3 were 9.06, 8.74, and 3.73 respectively, which together accounted for about 50% of the total variance. Factor 1 (“recognition”) loaded principally on the items about recognition, interaction, and control. Factor 2 (“extrinsic”) loaded principally on items about extrinsic rewards. Factor 3 (“growth”) loaded principally on items about personal growth (such as research and career advancement). All the 31 items of the MMSS could be covered by the predominant loading by any one of these three factors.

Correlation between Mean McCloskey / Mueller Satisfaction Scale Subscores and Socio- demographic and Work-related Variables and Stress

For each of the three groups of MMSS items, which were respectively loaded predominantly by each of the three factors, a corresponding mean MMSS subscore (AMSRECON for Factor 1, AMSEXTF for Factor 2, and AMSGROWF for Factor 3) was calculated. For each subject this entailed summing the scores of the MMSS items in each group and dividing by the number of items in the group (Table 6). The subjects were most satisfied with the group of items corresponding to recognition, interaction and control, followed by items corresponding to extrinsic rewards; they were dissatisfied with items corresponding to personal growth.

The mean AMSRECON subscore was found to show significant negative correlations with the overall mean score for the global ASQ (r = –0.21, p = 0.03), and the mean scores for the subscales of job-induced tension (r = –0.21, p = 0.01) and somatic tension (r = –0.24, p = 0.01). No significant association was found between AMSRECON and the mean score for the subscale of general fatigue and uneasiness, gender, marital status, number of children, education level, years of nursing experience, rank, and work setting. Multiple linear regression analysis showed that the mean score for the somatic tension subscale (β = –0.27, p = 0.01) remained a significant predictor for AMSRECON. Thus, having less somatic tension predicted greater job satisfaction in the area of recognition, interaction, and control.

The mean AMSEXTF subscore was found to have significant negative correlations with the overall mean score for the ASQ (r = –0.36, p < 0.001), and the mean scores for all the three subscales, i.e. job-induced tension (r = –0.29, p = 0.003), somatic tension (r = –0.35, p < 0.001), and general fatigue and uneasiness (r = –0.33, p = 0.001). The mean AMSEXTF subscore was significantly higher among ward managers (mean ± SD, 3.69 ± 0.43) than the remainder (i.e. NOs, RNs, and ENs as a group; mean, 3.23; SD, 0.53) [t = 2.64, df = 100, p = 0.01]. The AMSEXTF subscore was also significantly higher in the group of CPNs (mean, 3.62; SD, 0.25) than in ward-based psychiatric nurses (working in acute, subacute, and chronic wards; mean, 3.24; SD, 0.55) [t = 2.07, df = 100, p = 0.04]. Multiple linear regression analysis showed that the mean score for the somatic tension subscale (β = –0.29, p = 0.002), the rank of being a ward manager (β = 0.25, p = 0.01) and being a CPN (β = 0.19, p = 0.04) remained significant predictors for AMSEXTF. Thus, having less somatic tension, being a ward manager, and being a CPN predicted greater job satisfaction in the area of extrinsic rewards.

The mean AMSGROWF subscore was found to have negative correlation with the mean score for the subscale of general fatigue and uneasiness (r = –0.20, p = 0.05). No association was found between AMSGROWF and all other variables, including: the overall mean score for the ASQ, the mean scores for the subscales of job-induced tension and somatic tension, rank, work setting, gender, marital status, number of children, and years of nursing experience. The linear regression model showed that having less general fatigue and uneasiness (β = –0.20, p = 0.05) predicted greater job satisfaction in the area of personal growth.

Discussion

This study investigated a large number of psychiatric nurses working in community and ward settings. On average, they were experienced (mean of about 14 years of nursing experience) and were of different ranks, ranging from ENs to ward managers. A significant proportion (> 40%) had attained higher qualifications such as Bachelor or Master Degrees, apart from the minimum professional qualifications required for a psychiatric nurse.

Using the MMSS, the results showed that the subjects in our study were generally satisfied with their job, although only slightly higher than the neutral point (mean, 3.24). They were satisfied in 6 of 8 subscales and in 22 of 31 items. They were most satisfied with their extrinsic rewards (the 3 component items: salary, vacation, and benefits package ranked 1st, 4th, and 8th among all the items). The second most satisfying aspect was co-workers (the 2 component items: nursing peers and physicians ranked 2nd and 10th respectively). They were most dissatisfied with the balance of family and work. To a large extent this was due to dissatisfaction with 2 of the component items: child-care facilities (ranked the last) and opportunity for part-time work (ranked 29th). The other ‘dissatisfied’ subscale was professional opportunities, of which the 4 component items concerned with opportunities for research and professional interaction ranked from 23rd to 28th.

As compared with the overall MMSS sum score (mean ± SD, 98.6 ± 17.6) in another study with Palestinian nurses,²² our subjects had a comparable but slightly higher overall MMSS sum score (mean ± SD, 100.5 ± 15.6). For the Palestinian nurses, the most satisfying items were responsibility (mean ± SD, 3.88 ± 0.89), relationship with physicians (3.84 ± 0.90), and nursing peers (3.81 ± 0.94); and the most dissatisfying were opportunity to write and publish (2.01 ± 1.10), nursing research (2.05 ± 1.13), and child-care facilities (2.12 ± 1.29). In a British study with nurses of an acute hospital, the subscales with the highest satisfaction were co-workers (3.8) and extrinsic rewards (3.5), and those with the greatest dissatisfaction were professional opportunities (2.6), and control and responsibility (2.7); the most satisfying items being vacation, hours of working, and nursing peers; and the most dissatisfying were child- care facilities, compensation for working weekends, and control over work conditions.²³ Our findings that the extrinsic rewards (including salary and vacation) and co- workers (including nursing peers and physicians) were most satisfying and that the balance of family and work (including child-care facilities and part-time work) and professional opportunities (including opportunities for research and publication) were most dissatisfying aspects therefore echoed the findings of previous studies. These findings suggest some general satisfying and dissatisfying elements for the nursing profession across different clinical specialties and different cultures. Extrinsic rewards such as salary and co-workers including nursing peers and physicians appear to be important satisfying elements, while the lack of professional opportunities for research and the lack of child-care facilities appear to be significant dissatisfying elements. The latter elements may clearly warrant improving for nurses’ job satisfaction. In Hong Kong, no child-care facilities are provided for nurses, and there is no opportunity for part-time work. An interesting finding in our study was that the satisfaction with the salary was the highest (mean score, 4.13); this being the only item with a mean score above 4 and clearly higher than the second-ranked item, nursing peers (mean score, 3.85). However, we did not know whether there was any difference between permanent and contract staff in their satisfaction with the salary.

In our study, all the measures of the ASQ including the global mean score and the mean subscores were negatively associated with the overall mean score of the MMSS. On regression analysis, the mean subscore for somatic tension remained the strongest predictor for overall job satisfaction. This confirmed the previous finding that job satisfaction was most strongly (and negatively) associated with stress.⁷

Another important significant predictor for overall job satisfaction in our study was being a ward or service manager. It was previously found that nurses’ job satisfaction was increasingly influenced by their position within the organisation and organisational culture.¹⁵ In this study, ward leaders were dissatisfied because of role conflict and strain, while nurses of lower clinical grades were concerned with managerial and resource limitations on their ability to provide good quality care. Similarly, it may be possible that our ward and service managers derived greater job satisfaction from their role as leaders and managers of their wards and services, especially in the present climate when effective management and efficient use of resources are very important given the limited resources available, while nurses of lower clinical grades might be concerned with managerial and resource limitations on providing good quality care. This may need further exploration.

One of the most important and interesting findings came from factor analysis. Instead of the 8 subscales identified by the original architects of the MMSS,²¹ we identified 3 factors: recognition, interaction and control (‘recognition’), extrinsic rewards (‘extrinsic’), and personal growth (‘growth’). Our ‘extrinsic’ factor covered predominantly the items of the subscales of extrinsic rewards, scheduling and the balance of family and work of the MMSS. This suggests that these 3 subscales as a whole may actually be considered as one factor. This is not surprising, as they all correspond to the original dimension of ‘safety’ of the 3 dimensions proposed by McCloskey,^21,39 and to the dimensions of physiological needs and safety of Maslow.^17,46 Our ‘recognition’ factor covered the items of the subscales of praise and recognition, co-workers and interaction, together with certain items concerned with control over one’s work. This corresponds mainly to the ‘social’ dimension of McCloskey and the need to belong and self-esteem proposed by Maslow.^17,21,39,46 Our ‘growth’ factor covered the items corresponding to the need for personal achievement and advancement, and corresponds to the need for self-actualisation of Maslow and the ‘psychological’ dimension of McCloskey.^17,21,39,46 It appears that our study has provided further support for Maslow’s theory and McCloskey’s 3 dimensions,^17,21,39,46 and suggests that these theories may be applicable to the job satisfaction of psychiatric nurses (Table 7). Although the previous Palestinian study²² identified 4 factors— interaction, extrinsic rewards, control over environment, and personal achievement—which together accounted for about 36% of the overall variance and loaded mainly on 26 items of the MMSS, a closer examination of the items yielded similarity between their findings and ours; their ‘interaction’ and ‘control over environment’ corresponded to our ‘recognition’ factor, their ‘extrinsic rewards’ corresponded to our ‘extrinsic rewards’ (except that ours covered all the items of the 3 subscales), and their ‘personal achievement’ corresponded to our ‘personal growth’.

Interestingly, measures of stress remained as associated factors or significant predictors for MMSS subscores. As for the MMSS subscore for the ‘extrinsic’ factor in our study, we found that being a ward manager or service manager and being a CPN were associated factors, and remained significant predictors, apart from the subscore for somatic tension in multiple regression analysis. Previous studies found that community nursing or community psychiatric nursing was more satisfying, although possibly more stressful, than nursing in hospital setting.^1,8,26 We did not find community psychiatric nursing to be significantly associated with overall job satisfaction. However, community psychiatric nursing was a significant predictor for the MMSS subscore for the ‘extrinsic’ factor only, even though the factor of stress was also included in the regression analysis. As the ‘extrinsic’ factor covered the items concerned with scheduling, one possible explanation for our finding could be that our CPNs generally worked during office hours, as opposed to the shift duties of ward- based nurses. Similarly, this may also explain why being a ward manager was also a significant predictor for the ‘extrinsic’ subscore; ward managers work slightly fewer night shifts and have more office-hour shifts compared to nurses in lower clinical grades. Ward managers also have a higher salary, which could well be relevant.

Our study focused on psychiatric nurses. A previous British study³ was conducted on a multidisciplinary sample of mental health staff, including psychiatrists, nurses, occupational therapists, social workers and psychologists. It found that being a psychiatrist and not being a social worker predicted greater job satisfaction. Another recent study with mental health social workers in England and Wales⁵ reported high levels of stress and low levels of job satisfaction among the respondents, though that study was limited by the measurement of job satisfaction with a single item only. It may also be worth conducting further studies on job satisfaction and stress among other disciplines of the mental health staff in Hong Kong.

A limitation of this study was our low response rate (about 50% only). This was lower than the rate of more than 80% in a British study using the MMSS,²³ but comparable to the 50% rate of another study employing the MMSS as well as an investigator-designed demographic questionnaire.²² A number of nurses on prolonged leave might not have been able to respond to our questionnaires. In addition, there were significantly more ward managers and fewer ENs in the group of respondents, which could have affected the statistical analysis involving rank. It may be argued that nurses who are more satisfied with their job are more likely to have responded to our questionnaires. Furthermore, non- responders might also have had a different perspective on job satisfaction.

An important issue in the use of the MMSS is the possible confusion between level of satisfaction and the importance of the item as judged by the subject.²² For example, regarding opportunities for research and publication, subjects could be confused when scoring the importance of as opposed to the satisfaction with these items.²² This issue may be better addressed by the type of questionnaire which asks for subjects’ level of satisfaction and also their perceived importance of different job aspects, such as the IWS which gives a weighted score for job satisfaction.^34,35 However, the latter type of questionnaire may be more complicated to use, and may deter responses.

It is possible that the results in this study reflect only the characteristics of the local working and organisational culture of our area. A larger sample involving psychiatric nurses of different hospitals could have enhanced the generalisability of the results, and increased the power of the study to detect significant differences and associations. However, given the limited resources available to us, the use of well-established rating instruments together with systematic collection and analysis of data can serve as a good starting point. Further studies in this field, extended to other mental health disciplines (especially in the local context), may facilitate recruitment and retention of mental health staff and improve patient care.^13-15

Acknowledgements

The authors would like to thank Prof JC McCloskey and Prof CW Mueller for their permission to use the McCloskey / Mueller Satisfaction Scale in the study; Prof TKS Wong, Ms FK Lee, and Ms ATY Shiu for sharing their invaluable research experience; and Dr Joseph Lee, Dr PC Iu, and the nurses of Castle Peak Hospital for their support and participation in the study.

References

1. Carson J, Leary J, de Villiers N, Fagin L, Radmall J. Stress in mental health nurses: comparison of ward and community staff. Br J Nurs 1995;4:579-82.
2. Hopkinson PJ, Carson J, Brown D, Fagin L, Bartlett H, Leary J. Occupational stress and community mental health nursing: what CPNs really said. J Adv Nurs 1998;27:707-12.
3. Prosser D, Johnson S, Kuipers E, Szmukler G, Bebbington P, Thornicroft G. Mental health, “burnout” and job satisfaction among hospital and community-based mental health staff. Br J Psychiatry 1996;169:334-7.
4. Prosser D, Johnson S, Kuipers E, Dunn G, Szmukler G, Reid Y, et al. Mental health, “burnout” and job satisfaction in a longitudinal study of mental health staff. Soc Psychiatry Psychiatr Epidemiol 1999;34:295- 300.
5. Evans S, Huxley P, Gately C, Webber M, Mears A, Pajak S, et al. Mental health, burnout and job satisfaction among mental health social workers in England and Wales. Br J Psychiatry 2006;188:75-80.
6. Rees D, Smith S. Work stress in occupational therapists assessed by the Occupational Stress Indicator. Br J Occup Ther 1991;54:289-94.
7. Blegen MA. Nurses’ job satisfaction: a meta-analysis of related variables. Nurs Res 1993;42:36-41.
8. Butterworth T, Jeacock J. Stress, coping, burnout and job satisfaction in British nurses: findings from the clinical supervision evaluation project. Stress Med 1999;15:27-33.
9. Cavanagh SJ. Nursing turnover: literature review and methodological critique. J Adv Nurs 1989;14:587-96.
10. Irvine DM, Evans MG. Job satisfaction and turnover among nurses: integrating research findings across studies. Nurs Res 1995;44:246- 53.
11. 1 Blegen MA, Mueller CW. Nurses’ job satisfaction: a longitudinal analysis. Res Nurs Health 1987;10:227-37.
12. Butler J, Parsons RJ. Hospital perceptions of job satisfaction. Nurs Manage 1989;20:45-8.
13. Weisman CS, Nathanson CA. Professional satisfaction and client outcomes. A comparative organizational analysis. Med Care 1985;23:1179-92.
14. Packard JS, Motowidlo SJ. Subjective stress, job satisfaction, and job performance of hospital nurses. Res Nurs Health 1987;10:253-61.
15. Tovey EJ, Adams AE. The changing nature of nurses’ job satisfaction: an exploration of sources of satisfaction in the 1990s. J Adv Nurs 1999;30:150-8.
16. Timmreck TC. Managing motivation and developing job satisfaction in the health care environment. Health Care Manag (Frederick) 2001;20:42-58.
17. Maslow AH. Motivation and personality. New York: Harper; 1954.
18. Herzberg F. The motivation-hygenic theory. In: Pugh D, editor. Organization theory. Harmondsworth: Penguin; 1966.
19. Koelbel PW, Fuller SG, Misener, TR. Job satisfaction of nurse practitioners: an analysis using Herzberg’s theory. Nurse Pract 1991;16:43,46-52,55-6.
20. Weiss DJ, Dawis RV, England GW, Lofquist LH. Manual for the Minnesota Satisfaction Questionnaire. Minneapolis: University of Minnesota, Work Adjustment Project, Industrial Relations Centre; 1967.
21. Mueller CW, McCloskey JC. Nurses’ job satisfaction: a proposed measure. Nurs Res 1990;39:113-7.
22. Abu Ajamieh AR, Misener T, Haddock KS, Gleaton JU. Job satisfaction correlates among Palestinian nurses in the West Bank. Int J Nurs Stud 1996;33:422-32.
23. Price M. Job satisfaction of registered nurses working in an acute hospital. Br J Nurs 2002;11:275-80.
24. Carr KK, Kazanowski MK. Factors affecting job satisfaction of nurses who work in long-term care. J Adv Nurs 1994;19:878-83.
25. Burnard P, Morrison P, Phillips C. Job satisfaction amongst nurses in an interim secure forensic unit in Wales. Aust N Z J Ment Health Nurs 1999;8:9-18.
26. Parahoo K. Job satisfaction of community psychiatric nurses in Northern Ireland. J Adv Nurs 1991;16:317-24.
27. Landeweerd JA, Boumans NP. Nurses’ work satisfaction and feelings of health and stress in three psychiatric departments. Int J Nurs Stud 1988;25:225-34.
28. Polit D, Hungler B. Essentials of research: methods, appraisals, and utilization. 4th ed. Philadelphia, Pennsylvania: Lippincott-Raven; 1997.
29. Waltz C, Strickland O, Lenz E. Measurement in nursing research. 2nd ed. Philadelphia, Pennsylvania: FA Davis; 1991.
30. McNeese-Smith DK. A content analysis of staff nurse descriptions of job satisfaction and dissatisfaction. J Adv Nurs 1999;29:1332-41.
31. Brayfield AH, Rothe HF. An index of job satisfaction. J Appl Psychol 1951;35:307-11.
32. Hackman JR, Oldham GR. Development of the job diagnostic survey. J Appl Psychol 1975;60:159-70.
33. Ironson GH, Smith PC, Brannick MT, et al. Construction of a job in general scale: a comparison of global, composite, and specific measures. J Appl Psychol 1989;74:193-200.
34. Stamps PL, Piedmonte EB. Nurses and work satisfaction: an index for measurement. Ann Arbor, Michigan: Health Administration Press Perspectives; 1986.
35. Stamps PL. Nurses and work satisfaction: an index for measurement. 2nd ed. Massachusetts: Amherst; 1997.
36. Traynor M, Wade B. The development of a measure of job satisfaction for use in monitoring the morale of community nurses in four trusts. J Adv Nurs 1993;18:127-36.
37. Yamashita M. Job satisfaction in Japanese nurses. J Adv Nurs 1995;22:158-64.
38. Allgood C, O’Rourke K, VanDerslice J, Hardy MA. Job satisfaction among nursing staff in a military health care facility. Mil Med 2000;165:757-61.
39. McCloskey J. Influence of rewards and incentives on staff nurse turnover rate. Nurs Res 1974;23:239-47.
40. Cheung YB, Shae WC, Wong TK, Luk CC, Fielding R. The development of a nurse’s job satisfaction measure in Hong Kong. Asian J Nurs Stud 1993;1:56-64.
41. Lee FK. Job satisfaction and autonomy of Hong Kong registered nurses. J Adv Nurs 1998;27:355-63.
42. Callaghan P, Shiu AT, Wyatt PA. Factors related to stress and coping among Chinese nurses in Hong Kong. J Adv Nurs 2000;31:1518-27.
43. Census and Statistics Department. 2001 population census. Hong Kong: Hong Kong SAR Government; 2001.
44. House RJ, Rizzo JR. Role conflict and ambiguity as critical variables in a model of organizational behaviour. Organiz Behav Human Perform 1972;7:467-505.
45. Cook JD, Hepworth SJ, Wall TD, Warr PB. The experience of work: a compendium and review of 249 measures and their use. London: Academic Press Inc. Ltd; 1981.
46. Maslow AH. A theory of human motivation. In: Motivation and personality. 2nd ed. New York: Harper & Row Publishers, Inc.; 1970.