Job satisfaction has been defined as the way people feel about their work and the different aspects of their jobs, and is the extent to which people like or dislike their jobs (Spector, 1997). It has also been suggested that job satisfaction is the pleasurable or positive emotional state resulting from the appraisal of one's job or job experiences (Locke, 1984). Job satisfaction in an organisation is considered an important determinant of work productivity and a good indicator of job quality. There is evidence that high levels of job satisfaction are associated with reduced levels of staff turnover, employee absenteeism and occupational accidents (European Foundation for the Improvement of Living and Working Conditions, 2007).
On review of the local literature there are many studies that have been conducted in Greek hospital settings relating to job satisfaction and motivation but involve nursing staff as a whole with no identification of a sub-group of midwives (Paleologlou et al, 2006; Kontodimopoulos et al, 2009). Job satisfaction and motivation of nurses was generally found to be moderate-to-low and depending on the study design, hospital setting and research instrument, there were several different determinants of job satisfaction, such as levels of autonomy, recognition, achievements and remuneration.
There have also been several studies evaluating job satisfaction in other countries involving mostly nursing staff and with few studies conducted on midwifery staff. However, most of these studies are now quite old. A study conducted in Ireland by Curtis (2007) revealed moderate levels of job satisfaction and factors that contributed most to nurses' job satisfaction were professional status, interaction and autonomy. In an earlier study in England by Lavender and Chapple (2004) autonomy was also highlighted to promote job satisfaction of midwives' working in maternity services. A meta-analysis of 31 studies on nurses' job satisfaction in the United States by Zangaro and Soeken (2007) identified three sets of main predictors: job stress, nurse-physician collaboration, and levels of autonomy. Two studies from Australia involving midwives showed that job satisfaction depended mostly on the positive interaction with women in their care (Watson et al, 1999; Sullivan et al, 2011). Global satisfaction was high when midwives felt they had made a difference to women and saw them happy. Another study of midwives' job satisfaction in the UK by Martin and Bull (2009) demonstrated that strict adherence to protocols and guidelines in order to optimise healthcare and prevent litigation had a detrimental impact on the global satisfaction index. It was considered that individualised and woman-centred care could not be delivered by midwives and they could not make use of their personal experience and resourceful thinking to cope with even the slightest challenges in everyday midwifery practice.
Obstetric care in Greece
Over the past decades, birth has become highly medicalised and almost all are performed in hospital settings in both the Greek National Health Service and the private health care system (Kontoyiannis and Katsetos, 2008). Unfortunately there are no primary health-care settings, no community midwives or midwifery-led birth units in the current Greek healthcare system, and home births facilitated by independent practising midwives account for less than 0.9% of total births (Kontoyiannis and Katsetos, 2008). Birth units in secondary or tertiary level general hospitals of the Greek National Health Service and in tertiary private hospitals have become the main setting for childbirth in Greece. Even though Greek midwives are professionally entitled to practise autonomously and perform deliveries independently, their role has been unfortunately restricted to assisting doctors during birth. In the provision of antenatal and postnatal care, midwives are part of a combined team with doctors but they are not the lead professionals even in cases of pregnant women with low-risk profiles. The autonomy of Greek midwives tends to be limited and there is much dependency on obstetricians in their everyday role, a fact that is also encountered elsewhere in the world when midwives practice in a hospital (Sato and Adachi, 2013).
Aim of the study
This prospective observational study was conducted to measure the job satisfaction of midwives who work in public and private hospitals in the capital city of Athens and to determine the main determinants of their global satisfaction. This is the first study investigating these issues in hospital-based midwifery care in Greece.
Research instrument
There have been many attempts in literature to explain what factors affect the levels of job satisfaction. According to theorist Herzberg et al (1966), job satisfaction was ultimately determined by two different sets of factors. The first set of factors, which represented the ‘motivation’ factors led to satisfaction when they were present in adequate quantities. The second set of factors, which represented the ‘retention’ factors caused dissatisfaction when they were deficient. Both motivation and retention factors determined job satisfaction; however, neither the absence of motivators necessarily resulted in job dissatisfaction nor the presence of retention factors necessarily resulted in high levels of job satisfaction.
The two-factor theory of Herzberg has extended the knowledge in literature regarding the factors that enhance motivation and prospectively work performance. In this study, we have used a validated questionnaire as primary research instrument that has been based on the motivation–retention factors of Herzberg's theory and was developed by Labiris et al (2008). The initial questionnaire presented satisfactory validity and reliability and was used for the measurement of job satisfaction in a mental health Greek NHS hospital. For the purposes of this study, the items in the questionnaire were modified to reflect midwifery practice issues and needs. Close-ended questions in the form of a five point Likert scale were used to avoid misinterpretation of results, in which 1 corresponded to ‘strongly agree’, 2 to ‘agree’, 3 to ‘neither agree or disagree’, 4 to ‘disagree’ and 5 to ‘strongly disagree’.
Motivation–retention
The motivation factors involved: ‘achievement’, ‘recognition’, ‘advancement’, ‘the work itself’, and the ‘possibility of growth and responsibility’. The retention factors involved: ‘salary’, ‘interpersonal relations’, ‘supervision’, ‘organisation, policy and administration’, and ‘working conditions and job security’. These motivation-retention factors formed the basis of the questions on the questionnaire.
Participants and ethical approval
Midwives employed in public and private hospitals in the capital city of Athens were invited to participate in the study. The only inclusion requirement to be met was for midwives to have worked in their current hospital for more than 6 months. Midwives were instructed to restrict their views in relation only to the current working environment and not to previous work institutions.
The questionnaire was administered to each participant by the researcher and was collected back in a sealed envelope. In order to avoid linkage of any data to a midwife, the sealed envelopes were opened at the end of the study period. The questionnaire was also accompanied by a letter explaining the objectives of the study and assuring that all data were anonymous and to be kept confidential and not be shown to any employer. The questionnaire was initially pilot-tested in a small sample of midwives (n=20) and was found to be easy and simple to complete and short in duration (approximately 10 minutes).
Written informed consent was obtained by each participant in the study. The study design and its objectives were approved by the Ethics Committee of the Faculty of Social Sciences at the Hellenic Open University in Greece.
Data analysis
Job satisfaction was considered the dependent variable, whereas the public or private nature of hospital settings, sociodemographic and job-related data and the motivation-retention factors of Herzberg's theory were the independent variables. The reliability and validity of the questionnaire was evaluated with estimation of the Cronbach α coefficient and the goodness of fit index (GFI) respectively.
For correlation analysis, job satisfaction was initially considered as a quantitative variable and Pearson's correlation coefficient (r) was calculated between job satisfaction index (GSI) and participants' demographics and the motivation and retention factors.
Job satisfaction was subsequently considered a qualitative variable and was considered as high if participants declared that they were ‘very satisfied/satisfied’, which represented items 1 and 2 in the five point Likert scale. Job satisfaction was considered as low if the participants declared that they were ‘neither satisfied/dissatisfied, dissatisfied, very dissatisfied’, which represented items 3, 4 and 5 in the five point Likert scale. Cohen effect size analysis was performed and the effect size was calculated between job satisfaction index (GSI) and Herzberg's motivation and retention factors.
Statistical analysis for quantitative data was performed with the use of SPSS statistical software version 17.0 (Chicago, IL, USA) and level of significance was set at P=0.05 (two-tailed).
Results
Questionnaire's reliability-validity
The questionnaire used in the study had adequate reliability and validity as a research instrument. Cronbach's α and GFI values were calculated. Cronbach's α values ranged from 0.50 to 0.81, while GFI values ranged from 0.91 to 0.97.
Demographic characteristics of midwives
Over the time period of the study (April–September 2012), 168 questionnaires were administered and 145 were completed and handed back in a sealed envelope (response rate=86.3%). The questionnaires were from four public hospitals (n=91) and five private hospitals (n=54) in the capital city of Athens, Greece. This represented 8.7% of midwives practising hospital-based midwifery in Athens, according to data provided by the Hellenic Society of Midwives for the years 2011 and 2012.
More than 95% of midwives were female with those working in private hospitals being younger than the midwives in public hospitals (36.5±7.8 vs 39.1±9.3 years; P=0.10). Midwives working in the public sector had more years of professional experience (12.4±9.9 versus 8.2±7.2 years; P=0.01) and were paid a much higher salary. Midwives working in public and private hospitals had the same rate of possession of post-graduate degree (approximately 1 in 10), and had the same family status (1 in 2 were married) and number of children (median: 1 child).
Job satisfaction measurement
The midwives that were ‘very satisfied/satisfied’ in the questionnaire, represented 45.5% (66/145) of the total sample, whereas in the sub-group analysis they represented 50.5% (46/91) and 37.1% (20/54) in the public and private sector, respectively (P=0.12). Mean scores for job satisfaction with use of the five point Likert scale were 2.74±0.90, 2.74±0.98 and 2.76±0.75 in the total sample, public hospital and private hospital, respectively (P=0.88). This finding indicates that job satisfaction was similar between public and private hospitals.
Association between job satisfaction and midwives' demographics
Pearson correlation analysis between job satisfaction and sociodemographic and job-related data revealed that in the total cohort there was a significant correlation between job satisfaction and age, working status and educational level. The younger midwives that had just entered the midwifery practice were more satisfied than those who were older and those having worked in the service for more than a year. In contrast, the permanent members of staff and those of higher educational level were less satisfied than those with a temporary term contract and those without a postgraduate degree. In the public sector, the permanent members of staff and those of higher educational level were significantly less satisfied than those with a temporary term contract and those without a postgraduate degree. Finally, in the private sector those who were working with temporary term contracts were significantly more satisfied than the permanent members of staff.
Association between job satisfaction and motivation–retention factors
Table 1 presents the Pearson correlation analysis between job satisfaction and Herzberg's theory factors. It is noted that ‘recognition’ and ‘organisation, policy and administration’ demonstrated the strongest correlation with job satisfaction. Tables 2 and 3 present the Cohen effect size between job satisfaction and Herzberg's theory motivation and retention factors. Again, ‘recognition’ demonstrated the strongest effect on high job satisfaction in all sub-groups, whereas ‘organisation, policy and administration’ demonstrated the strongest effect on high job satisfaction in the total sample and in public hospitals. Tables 1, 2 and 3 demonstrate that the retention factors had higher Pearson correlation coefficients and greater Cohen effect size coefficients in comparison to the motivation factors of Herzberg's theory. This means that they contributed to a greater extent to the job satisfaction index of midwives when compared to the motivation factors.
| Job satisfaction | |||||
|---|---|---|---|---|---|
| Total sample (n=145) | Public hospitals (n=91) | Private hospitals (n=54) | |||
| Recognition | 0.612b | Recognition | 0.646b | Recognition | 0.513b |
| Advancement | 0.502b | Advancement | 0.523b | Advancement | 0.448b |
| Work itself | 0.414b | Work itself | 0.411b | Work itself | 0.459b |
| Achievement | 0.398b | Achievement | 0.407b | Achievement | 0.377b |
| Responsibility | 0.363b | Responsibility | 0.319b | Responsibility | 0.467b |
| Organisation policy and administration | 0.670b | Organisation policy and administration | 0.712b | Organisation policy and administration | 0.589b |
| Supervision | 0.629b | Supervision | 0.649b | Supervision | 0.580b |
| Working conditions and job security | 0.584b | Working conditions and job security | 0.644b | Working conditions and job security | 0.459b |
| Interpersonal relations | 0.572b | Interpersonal relations | 0.581b | Interpersonal relations | 0.754b |
| Salary | 0.506b | Salary | 0.546b | Salary | 0.468b |
Correlation significant at P=0.05
Correlation significant at P=0.01
| Very dissatisfied/Dissatisfied/Neither satisfied, nor dissatisfied (Mean±SD) | Very satisfied/Satisfied (Mean ±SD) | P-value Student's t-test | Effect size | |
|---|---|---|---|---|
| Total sample (n=145) | ||||
| Recognition | 3.8±1 | 2.7±1 | <0.001 | 1.14 |
| Advancement | 4.5±0.7 | 3.6±1.1 | <0.001 | 0.93 |
| Work itself | 2.3±0.7 | 1.7±0.6 | <0.001 | 0.90 |
| Achievement | 3.9±1 | 3±1 | <0.001 | 0.89 |
| Responsibility | 3.1±0.9 | 2.6±1 | <0.001 | 0.60 |
| Public hospitals (n=91) | ||||
| Recognition | 3.8±1.1 | 2.6±1 | <0.001 | 1.14 |
| Advancement | 4.5±0.8 | 3.6±1.2 | <0.001 | 1.01 |
| Achievement | 3.9±1.1 | 2.9±1 | <0.001 | 0.95 |
| Work itself | 2.2±0.8 | 1.6±0.5 | <0.001 | 0.79 |
| Responsibility | 3.1±0.9 | 2.6±0.9 | 0.019 | 0.50 |
| Private hospitals (n=54) | ||||
| Recognition | 3.8±0.8 | 2.9±0.9 | <0.001 | 1.07 |
| Work itself | 2.5±0.6 | 1.9±0.6 | 0.001 | 0.97 |
| Advancement | 4.4±0.7 | 3.8±1.1 | 0.008 | 0.74 |
| Responsibility | 3.2±0.9 | 2.5±1.1 | 0.012 | 0.73 |
| Achievement | 3.8±0.9 | 3.1±1 | 0.012 | 0.72 |
| Very dissatisfied/Dissatisfied/Neither satisfied, nor dissatisfied (Mean±SD) | Very satisfied/Satisfied (Mean ±SD) | P-value Student's t-test | Effect size | |
|---|---|---|---|---|
| Total sample (n=145) | ||||
| Organisation policy and administration | 3.4±0.7 | 2.5±0.6 | <0.001 | 1.36 |
| Working conditions and job security | 3.4±0.6 | 2.5±0.8 | <0.001 | 1.22 |
| Interpersonal relations | 3.2±0.6 | 2.5±0.6 | <0.001 | 1.22 |
| Supervision | 3.5±0.9 | 2.5±0.8 | <0.001 | 1.20 |
| Salary | 3.7±0.7 | 2.9±0.8 | <0.001 | 1.05 |
| Public hospitals (n=91) | ||||
| Organisation policy and administration | 3.5±0.7 | 2.4±0.6 | <0.001 | 1.55 |
| Working conditions and job security | 3.4±0.7 | 2.3±0.7 | <0.001 | 1.42 |
| Supervision | 3.7±0.9 | 2.5±0.8 | <0.001 | 1.33 |
| Interpersonal relations | 3.2±0.6 | 2.5±0.6 | <0.001 | 1.19 |
| Salary | 3.6±0.7 | 2.8±0.8 | <0.001 | 1.13 |
| Private hospitals (n=54) | ||||
| Interpersonal relations | 3.3±0.6 | 2.6±0.6 | <0.001 | 1.21 |
| Supervision | 3.4±0.9 | 2.5±0.7 | 0.001 | 1.03 |
| Organisation policy and administration | 3.4±0.7 | 2.6±0.7 | 0.001 | 1.02 |
| Salary | 3.9±0.8 | 3.2±0.7 | 0.005 | 0.83 |
| Working conditions and job security | 3.4±0.6 | 2.9±0.8 | 0.007 | 0.76 |
Logistic regression analysis was also performed with job satisfaction as the dependent variable and the public or private nature of hospital settings, the sociodemographic and job-related data and the motivation-retention factors as the independent variables. In the total cohort, it was found that main predictors of job satisfaction was ‘work itself’, ‘supervision’, and ‘salary’. In the public sector, main predictors was ‘work itself’ and ‘supervision’. In the private sector, the main predictors of job satisfaction were ‘interpersonal relations’. The public or private nature of the hospital had no significant effect on the levels of job satisfaction.
Discussion
One of the primary outcomes of the study was to determine the global satisfaction index in hospital-based midwifery practice. The midwives participating in the study presented with average scores in the global satisfaction index and overall 45.5% reported being ‘satisfied/very satisfied’. Comparable rates of job satisfaction have been reported by Aiken et al (2001) in nursing staff in other countries, such as 41% in the US, 38% in Scotland and 36% in England. Kontodimopoulos et al (2009) also reported on low rates of nurses' job satisfaction in Greek NHS hospitals. Unfortunately in all the above described studies, data for midwifery practice are absent or missing and therefore comparisons are made against nursing staff.
According to the results of the correlation analysis there was a significant correlation in the total cohort between job satisfaction and age, working status and educational level. Younger midwives that had just entered the midwifery practice were more satisfied than the older midwives. In contrast, the permanent members of staff and those of higher educational level were less satisfied than those with a temporary term contract and those without a postgraduate degree. In the public sector, the permanent members of staff and those of higher educational level were significantly less satisfied than those with a temporary term contract and those without a postgraduate degree. Similarly, in the private sector, the permanent members of staff were significantly less satisfied than those working with temporary term contracts. The latter finding can be explained by the fact that permanent members of staff were older in age and therefore more likely to be less satisfied. This finding of increasing job satisfaction with younger age has been indicated in some surveys, while others indicate the opposite (European Foundation for the Improvement of Living and Working Conditions, 2007). In other surveys, no clear association has been shown between age and job satisfaction (European Foundation for the Improvement of Living and Working Conditions, 2007).
Another significant finding of the study was that the retention factors had higher Pearson correlation coefficients and greater Cohen effect size coefficients in comparison to the motivation factors of Herzberg's theory. This means that they contributed to a greater extent to midwives' job satisfaction when compared to the motivation factors. There are several studies in literature that have reached to similar conclusions. Pugh et al (2013) indicated that the ‘working conditions’ factor is a primary determinant of job satisfaction. Moreover, essential determinants of job satisfaction have been demonstrated for the ‘interpersonal relations’ (Hendel et al, 2007), ‘supervision’ (Begat et al, 2005), and ‘salary’ (French et al, 2007). Among the motivation factors in this study, it was noted that ‘recognition’ demonstrated the strongest correlation with job satisfaction and the strongest effect on ‘high’ job satisfaction in all sub-groups. This is consistent with other studies that have confirmed that ‘recognition’ is a well known contributor of motivation and prospectively to the global satisfaction (Krogstad et al, 2006; Labiris et al, 2008).
Logistic regression analysis identified ‘work itself’, ‘supervision’ and ‘salary’ as main predictors of job satisfaction in the total cohort of midwives. In the public sector, main predictors were ‘work itself’ and ‘supervision’, whereas in the private sector ‘interpersonal relations’ were the main predictors of job satisfaction. The public or private nature of the hospital made no difference on the levels of job satisfaction. The motivation factor of ‘work itself’ has been reported by Kontodimopoulos et al (2009) to be an important factor affecting job satisfaction. The retention factor of ‘supervision’ reflects the relationship between health professionals and their senior leaders and has also been indicated as a significant determinant of job satisfaction (Begat et al, 2005). The retention factor of ‘interpersonal relations’ has also been indicated in other studies to be a significant predictor of global satisfaction in both private and public healthcare systems (Rosenstein and Daniel, 2005; Kontodimopoulos et al, 2009).
The retention factor ‘salary’ has been shown to be a main predictor of job satisfaction in other studies (Kontodimopoulos et al, 2009). However, Spector (1997) reported that ‘salary’ had no clear effect on job satisfaction. In this study, the finding that ‘salary’ was a significant determinant of midwives' job satisfaction may simply reflect the national economic crisis the country has been experiencing over the past years.
Limitations
This study involved midwives practising in the capital city of Athens. If participants were to be included from other geographical regions, the results may have been different thus reflecting the diversity of working environments depending on the local health institutes. The impression was also given that job satisfaction was dependent on the work intensity levels and the length of time interval for rotation between different departments and clinics. Midwives reported low job satisfaction if they were not transferred to other clinics with less work intensity such as the outpatient's clinic within a certain time interval. Unfortunately, the study was underpowered to investigate the correlation between job satisfaction and department or clinic. Another limitation involved the use of a Likert scale with the associated shortcomings of using such a scale. The Likert scale is a unidimensional scale and gave the participants of our study only five options of choice. In reality, however, the attitudes of respondents for an item exist on a multidimensional continuum and therefore our study may have fallen short of measuring their true attitudes. Also, the questionnaire did not include any free-text boxes for comments that could have added valuable qualitative data to our study. Nevertheless, personal comments verbally reported from midwives on completion of the questionnaire were taken in consideration by the researcher but no results could be produced as this was not included in the initial design of the study.
Conclusion
Our study demonstrated that 45.5% of participants were ‘satisfied/very satisfied’, which was comparable with rates reported in literature. ‘Recognition’ presented the strongest correlation with job satisfaction and the strongest effect on ‘high’ job satisfaction in all sub-groups. Furthermore, main predictors of job satisfaction in the total cohort of midwives were ‘work itself’, ‘supervision’ and ‘salary’. In the Greek public healthcare system main predictors were ‘work itself’ and ‘supervision’, whereas in the private healthcare system ‘interpersonal relations’ were the main predictors of job satisfaction. In addition, the public or private nature of the hospital settings had no significant effect on midwives' job satisfaction.
In order to design strategies in the Greek healthcare system that will promote midwives' job satisfaction, further studies are needed with greater sample sizes so as to increase the generalisability of our findings. Certain targeted measures could be implemented to improve midwives' job satisfaction such as job enrichment plans for the factor of ‘work itself’. Moreover, modifications in the feedback process could be considered for the factor of ‘supervision’. Initiatives to promote social skills and interpersonal relations among staff could be undertaken at a central policy level to address the factor of ‘interpersonal relations’, whereas the provision of fringe benefits could be considered for the retention factor of ‘salary’.