The current financial crisis has heavily impacted the public sector in Europe and, to a greater extent, in countries in which the crisis was more severe and in those where the public system has higher costs. According to the WHO, “the global financial crisis that began in 2007 can be classified as a health system shock” [1].

In this context, Italy is an interesting case study. Italy has a widespread national public health system, in which, the need for reducing the high level of public deficit and debt led to the so-called “block of personnel turnover” policies, limiting Italian public sector in hiring new workers. At the same time, another consequence of the financial crisis is the continuous raising by law of the eligibility age for pension. Therefore, everyone who has a chance (in terms of age) to retire, they do it to avoid a change that could hamper their opportunity window. In fact, workers are threatened with a possible future political reform of the age for retirement, thus they retire even if they feel able to continue to work.

The result of this financial crisis is that Italy is facing a significant aging process in its workforce. Workers are forced to stay longer at work than they expected or desired. At the same time, workers from the public health sector are highly exposed to a number of job stressors, ranging from work overload, time pressures, low support and lack of role clarity. Research literature suggests that burnout, depression, and a low mental quality of life (QOL) are common among health care workers [2-7]. Further evidence showed that employee’s mental health may affect significantly quality of care, job satisfaction, intent-to-leave and increase the risk of suicidal ideation [8-14]. The public health care workers of the university system live the crisis to a greater extent. The health care workers who teach or operate in the medical faculties of public universities in Italy, unlike other European countries, receive only one salary for two jobs. They really should use a proportion of their working time for a job (treating patients) and a proportion for the other (teaching and research or support teaching and research). The present crisis in the two sectors can double, in this case, its impact on workers.

The main purpose of this study was to identify employees’ depression risk, burnout levels and quality of life, in a sample of workers of an Italian university agency (that manage two University Hospitals). The results were compared with data from community samples. Within the sample of workers, we analyzed the risk of burnout and the socio-demographic determinants and the associated factors related to working conditions (work history, type of department, type of work).

Table 1 shows the characteristics of the interviewed sample. Overall, 522 workers accepted to take part in the study, representing a 78% response rate (669 individuals).

There were 147 subjects (women: n = 87; 59%) who refused to participate or were not tracked because on vacation or have leaved work for long sickness. The distribution by age and sex of those who were not interviewed did not differ from the one of those who agreed to participate.

Women were 66.3% of the interviewed samples. Most were 40 years old or older (64.8%). Graduates represent around a half of the sample as did married / living with a partner and those who didn't have children.

The table also shows the time spent by subjects to complete the questionnaires and to answer to the interview: the 48.9% of the sample took less than 15 minutes, with a mean in the sample of 14.7 ± 8.9 minutes.

Table 1 illustrates the working employment contract of the sample. A significant portion of the sample (24.9%) work under a fixed-term contract, the vast majority is employed full-time (95.8%) and only a small portion was administrative staff. Only 36.3% earn 2,500 euros or more, and more than 40% worked from more than 20 years.

Table 2 compares the demographic variables of the interviewed sample with those of the two samples of the general population that serve as “normative” standard. The sample of this research is not representative of the adult population, and it is in fact a sample of older on average workers, almost 65% has more than 40 years and 33% have more than 50 (Table 2).

In addition, women were over-represented compared to the community samples.

A first confrontation without balancing highlights a frightening state of malaise in the study sample (Table 2).

The frequency of depressive symptoms (positive at PHQ-9) is indeed the 33.2% in the examined health workers against the 13.1% of the community sample (OR = 4.60; 95% CI 3.57-5.92; P <0.0001), while the frequency of individuals who accuse a low level of quality of life (below the national average minus one standard error) was 42.7% against 23.6% of the general population (OR = 2.40; Cl95% 1.96-2.94; P <0.0001).

The comparison made after indirect standardization by age and sex of the two normative samples confirms the state of malaise in the sample of workers with risk of low quality of life (47% in the study sample versus 25.9% in the “normative” sample, OR = 2.18; Cl95% 1.78-2.67) and positivity screened for depression (33.2% against 14.1% in the community sample, OR = 2.89; Cl95%).

Table 3 examines the factors associated with having a depressive episode (i.e., to be positive on the PHQ-9). To be male, and in youthful age were protective factors.

Those who were older and working in the University Hospitals had a risk that was not the simple sum of the two risk factors but they were amplified each other.

As for the risk related to the work in specific departments the “surgery” department only was associated with a higher risk, while working in laboratories was a protective factor.

Table 4 leads to the same type of analysis by professions: all workers of the university hospitals of all professions show a risk of developing a depressive episode higher than the regional normative sample except for technical and administrative staff.

Table 5 analyzes the quality of life in our sample that is combined for analysis at the national normative sample. The analysis was conducted using which cut-off the average of the national normative sample less a standard error calculated for a sample of the size of the one examined.

Males and less of 40 years old workers of both sexes showed a better quality of life. Also, examining work at the company is not a factor with respect to the risk of interaction by gender and age. Being a woman, more than 40 years old and working at the health care agency are risk factors that add without interaction (not amplify each other).

As shown in Table 6, is associated with a low quality of working life in the surgical wards; in Cardiology, Ophthalmology, Oncology (P <0.001) and Dermatology (P <0.05).

Table 7 shows the results concerning the risk of burn-out within the sample of this survey. The analysis has been conducted by logistic regression multivariate analysis with, as dependent variable, being positive to at least two of the three factors of the MBI questionnaire and as independent variables the demographic conditions (age and sex), the work history and work in specific departments. The work in the surgical wards is the only factor found to expose to the risk.

Table 8 analyzes the antecedents of burnout (positivity to at least two scales of MBI), subtracting in the multivariate analysis the kind of department / ward as an independent variable and introducing the type of work task and positivity at PHQ-9 and SF-12 (with the same cut-off already utilized). The Health Assistants have found to be protected from this risk. Also, the results highlights that score at SF-12 was negatively related to the risk of burnout (the highest score on this scale indicates a good level of quality of life) while in contrast the score of 8 or less at PHQ-9 is directly proportional. i.e., as better the quality of life, as lower the risk of burnout and, conversely, as higher the levels of depression, as higher the risk of burnout.

The present study found that in the overall health care workers sample, the frequency of positivity at the screener for Major Depressive Disorder is more than double of levels in the standardized community sample (33.3% vs 14.1%, p<0.0001). All the professionals, except the administrative staff and technicians (i.e. those who do not have contact with patients), showed a statistically higher frequency of individuals screened positive for depressive episode compared to the representative sample of the population of the region of Sardinia. Furthermore, being older than 40 years is not a risk factor for the normative regional control sample, but an interaction was found between being older than 40 years and working in the studied agency, with an amplification of the effect risk between the two conditions (age and working in the university hospitals). Results are in line with previous studies [25], which highlighted the protective factors of being a male and in youthful age. Furthermore, considering the specific work environment, results showed that working in surgery department is associated with a higher risk, while working in laboratories is a protective factor. These results are in line with previous researches that showed how distress, including anxiety, depression, alcoholism, substance abuse, are more prevalent among surgeons [27].

Past studies showed that being positive to the PHQ-9 is associated to a depression diagnosis. It is estimated that between 30% and 50% of positives to the PHQ-9 are clinically depressed [28, 29]. Considering our results, we found that positivity to the screening is very high among public health workers population (rate of 33%) while among the general population it is 13%. Taking into account that positivity to PHQ-9 corresponds to a clinical diagnosis of depression in about one third/fourth of cases, prevalence rates of clinical depression among public health workers and general population in these samples can be estimated respectively 8% and 4%.

Almost all of the professionals who carry out their work at the two university hospitals of a public Italian university care agency showed a risk of developing a depressive episode higher than the regional normative, and reported low level of perceived quality of life. These frequencies are higher, with statistically significant differences, with respect to a national representative sample of the population even after standardization by age and sex. The only exceptions are represented by the administrative staff and by technicians. Depression is a common mental disorder, with a prevalence of 14.6% among adults in high-income countries and 11.1% in developing countries [30]. Our results are in line with many studies that showed that the prevalence of depressive symptoms among health-care workers ranged from 18% to 41%.

Finally, considering the quality of working life, results are in line with the research literature. Health care workers are constantly exposed to the risk of low quality of working life. We did no find any difference when we considered profession and working unit. The quality of working life is considered as a factor beyond job satisfaction and is related to personnel’s well-being [14, 31]. In this sense, international studies showed how health-care workers (mainly nurses and physicians) were low in their quality of working life. Researchers mentioned that the health-care working context is inherently stressful [32], and many of the health-care workers are exhausted [33, 34].

When we considered job burnout levels, we found high burnout levels in the surgery unit. Results confirmed that incidence of burnout among health care workers affects more than the 30% of workers globally [35], reaching rates between 25% and 75% in some clinical specialties [36]. Specifically, among surgeons it has been reported reaching rates of 28% to 42% [27, 37, 38]. Burnout is very common among physicians if compared with depression and substance abuse. Considered as a clinical syndrome, burnout is characterized by emotional exhaustion, depersonalization, and a decreased sense of personal accomplishment. Surgeons’ burnout is linked to patient’s safety and quality of patient’s care, and contributes to medical errors [8, 9].

The highest frequency of positive screening for depression and lower quality of working life registered in the surgeon unit is of particular concern. Surgical practice is characterized by hard work, working long hours, dealing regularly with life-and-death environment with patients, and sometimes the sacrifice of personal life to the practice in the field [26]. According to the research literature, to reduce surgeon’s distress and then the risk of burnout, intervention programs should be aimed at reducing worker’s experience of stressors such as reducing workers’ workload, increasing their sense of control and promoting organizational health.

In sum, this study aimed to make a contribution towards public sector management, mostly considering the central role played by workplace characteristics that can improve employee well-being in public hospitals.

The present study has some limitations. First, a convenience sample has been used. This can limit the generalizability of the results, reducing external validity of the study. Another limitation is represented by the use of a self-reported questionnaire, which may yield a bias related to social desirability and common method bias. Finally, this study includes a cross-sectional design type and we are unable to examine the causal effect of the relationship between variables. This effect would be better analyzed through longitudinal studies, which would add something more about the development of mental health in working context.