What are the variables that affect performance and the level of job satisfaction?

PLoS One. 2018; 13[10]: e0205963.

Marie-Josée Fleury, Conceptualization, Data curation, Formal analysis, Funding acquisition, Methodology, Project administration, Writing – original draft, Writing – review & editing,1,2,* Guy Grenier, Conceptualization, Writing – original draft, Writing – review & editing,2 Jean-Marie Bamvita, Conceptualization, Formal analysis, Methodology, Writing – original draft, Writing – review & editing,2 and Lambert Farand, Conceptualization, Data curation, Writing – review & editing3

Kim Brimhall, Editor

Abstract

Recent mental health [MH] reforms have had a sharp impact on practices among MH professionals. A deeper understanding of factors contributing to their job satisfaction, in this context, may help improve quality and continuity of care. The purpose of this study was to identify variables associated with job satisfaction for 315 MH professionals in Quebec [Canada] after implementation of wide-ranging MH reforms. Job satisfaction was measured with the Job Satisfaction Survey. Independent variables were conceptualized within five domains: 1] Professional Characteristics, 2] Team Attributes, 3] Team Processes, 4] Team Emergent States, and 5] Organizational Culture. Univariate, bivariate and multivariate analyses were performed. Job satisfaction was significantly associated with absence of team conflict, stronger team support, better team collaboration, greater member involvement in the decision-making process [Team Processes], Affective commitment toward the team [Team Emergent States], as well as lack of a market/rational culture [Organizational Culture]. Job satisfaction was strongly related to team processes and, to a lesser extent, team emergent states.

Introduction

The renewal of mental health [MH] systems, with a goal of improving quality and continuity of care, has been on the agenda in several countries since the 2000s. The province of Quebec [Canada] exemplifies this international trend: the 2005 Quebec MH reform aimed to strengthen primary care MH services, and improve collaboration among psychiatrists and general practitioners through shared care initiatives in order to improve the integration of primary care and specialized MH services [1]. These changes had considerable impact on the practices of MH professionals. Those transferred from specialized services to primary care settings faced increased resource scarcity, for example. Asking professionals with different values, experiences, and practices to work together may have led to friction. While client satisfaction was the central focus of MH reform [2], job satisfaction among MH professionals was, and remains, a major consideration.

Job satisfaction, defined as “a pleasurable or positive emotional state resulting from the appraisal of one’s job or job experiences” [3], constitutes a key outcome of successful teamwork in the health field [4, 5]. Professionals who experience greater work satisfaction are better motivated to work together as a team, and collaborate with other groups or organizations. Job satisfaction also correlates with better quality of care [6]. A happy worker is more likely to resolve problems and make better decisions [7], which may, in turn, minimize or prevent errors [8]. Moreover, health care professionals whose services have a positive impact on clients experience greater satisfaction [9]. Job satisfaction also tends to reduce absenteeism [10], staff turnover [4], and burnout [11], deterring substance use and mental or physical distress [11] while contributing to staff dependability. By contrast, too many unsatisfied professionals may pose a serious barrier to successful healthcare reform [12, 13]. Dissatisfied professionals express more negative feelings toward their clients [14], provide less adequate service, and tend to terminate employment prematurely, with adverse effects for both service continuity and client MH recovery.

Job satisfaction is related to several categories of variables associated with the individual characteristics of professionals, expectations regarding workplaces and organizations, as well as relationships with other professionals [15, 16]. The prevailing organizational culture, described as the foundational assumptions, values and beliefs that underpin an organization [17] and are upheld by each of its members [18], also influences job satisfaction among professionals [5]. Several conceptual models have classified variables related to job satisfaction. The motivation-hygiene theory [19], for instance, distinguishes between variables related to job satisfaction, and to job dissatisfaction. While factors intrinsically related to the job [e.g. recognition, opportunities for personal achievement, work challenges] influence job satisfaction, other, more extrinsic, factors [e.g. interpersonal relationships, salaries] may be associated with job dissatisfaction [20]. According to the Input-Mediator-Outcomes-Input [IMOI] Model, individual characteristics [e.g. age, profession] are nested within team attributes [e.g. type of team, size] and, in turn, within the organizational context [e.g. organizational culture]. These variables, considered as inputs, influence two kinds of mediators: team processes, or the methods adopted by team members to work together and accomplish tasks [e.g. ways of dealing with conflict on teams] and team emergent states, or the outputs of teamwork, including motivation, cognition, and emotions such as trust [21]. Team processes and emergent states both generate job satisfaction [22], among other outcomes. Similarly, the heuristic Integrated Team Effectiveness Model [ITEM] states that task design variables [e.g. type and composition of team] are influenced by external environments including organizational context, culture, and the social or policy context, which teams manipulate, in turn [23]; potentially impacting team effectiveness, or the ability to reach objectives [24]. According to studies of job satisfaction based on the IMOI [22] and ITEM [23] models, associated variables may be grouped within five domains: 1] Professional Characteristics, 2] Team Attributes, 3] Team Processes, 4] Team Emergent States, and 5] Organizational Culture.

Professional Characteristics: Job satisfaction has been studied in terms of individual variables such as age, gender, type of profession, and seniority. For example, job satisfaction among physicians was found to be usually high [25] as opposed to nurses [26], social workers [25, 27], and other health care professionals [4]. High satisfaction among physicians may be due to perceptions that their tasks are relatively complex and of great importance [28]. As well, physicians enjoy high status in the professional hierarchy [29]. By contrast, younger professionals and those with less professional seniority have generally reported lower job satisfaction [12]. Significant associations between job satisfaction and gender have not yet been identified.

Team Attributes: These included group composition that may influence team effectiveness [30] and job satisfaction, more indirectly, due to the presence of professionals with different values and practices. Staff shortages on teams may also hinder job satisfaction [26], due to increased tasks and heavier caseloads. A comparative study of job satisfaction among psychiatric nurses [16] found the greatest satisfaction among those working in forensic services, which suggests that the type of clientele may influence professional job satisfaction. Moreover, job satisfaction was higher in smaller units where patients had less severe illnesses [31]. MH professionals dealing with aggressive patients, or those with severe MH and substance use disorders experienced high levels of stress [25].

Team Processes: Studies have investigated team processes extensively in relation to job satisfaction [25]. According to a literature review on job satisfaction, stress and burnout, positive collaboration among professionals, and frequent contact, were predominant factors in job satisfaction among community MH workers [4, 25, 32]. Job satisfaction also correlated with team and organizational support [26, 33], participation in decision-making [16, 26], team autonomy [25, 28], absence of conflict [16], and self-efficacy, i.e. “one’s belief [in his/her] ability and capacity to accomplish a task or cope with environmental demands” [34]. Furthermore, as knowledge related strongly to competent teamwork [35], there is a likely association between knowledge production and job satisfaction.

Team Emergent States: Variables associated with job satisfactions among team emergent states included team climate [16, 26, 36], trust, and affective commitment toward the team. [16, 36] Yet the association between job satisfaction and belief in the advantages of multidisciplinary collaboration has apparently not been assessed. The importance of this variable for interdisciplinary collaboration [37], and for job satisfaction, justifies further investigation.

Organizational Culture: Organizational culture is distinct from team climate, which refers to attitudes, norms and expectation in a specific context, and is also associated with job satisfaction [38]. Job satisfaction varies according to the prevailing corporate culture [5, 39], which is often defined along two axes: flexibility/stability in approaches to work, and internal/external focus of the governing structure [40]. The dominant organizational cultures fall into the following categories: a] family/clan culture [flexibility-internal focus], b] entepreunarial culture [flexibility-external focus], c] market/rational culture [stability-external focus], or d] hierarchical culture [stability-internal focus] [18]. Each culture highlights particular values: loyalty, development, participation and staff empowerment in the family/clan culture; listening capacity with clients, innovation and risk-taking in the entrepreneurial culture; competition, results-orientation, and achievement of measurable goals in the market/rational culture; and coordination, formalization, stability and efficiency in the hierarchical culture [18, 41]. Job satisfaction is expected to be higher in a family/clan culture, moderate in an entrepreneurial or market/rational culture, and lower in a hierarchical culture [42].

Overall job satisfaction in the health field has been the subject of various studies, mainly focused on specific groups of professionals such as nurses [15, 20, 43–45], social workers [27] or physicians [11]. Other research has assessed job satisfaction among health workers within specific settings such as community MH teams [25, 46] and acute care hospitals [4], or at particular stages of professional life [e.g. early career] [14]. No known studies have assessed job satisfaction among various types of MH professionals working in different settings [e.g. primary care, specialized MH services], however, or across different local health service networks. The new primary care MH teams developed in the context of the Quebec reforms often included professionals transferred from specialized MH services. where they had enjoyed considerable organizational and material support [33]. The shift of these MH professionals to primary care may have created some degree of job dissatisfaction, which warrants investigation.

A deeper understanding of the determinants of job satisfaction among MH professionals in the Quebec context, might contribute to better quality and continuity of care for service users. As such, this study aimed to identify variables linked to job satisfaction among 315 MH professionals in Quebec [Canada] following the implementation of a major MH reform. Based on a conceptual model adapted from the IMOI Model [22] and ITEM [23] as well as the literature reviewed above, we hypothesized that job satisfaction would more likely be associated with 1] Team Processes and 2] Team Emergent States than with Professional Characteristics, Team Attributes or Organizational Culture.

Materials and methods

Study design and data collection

This cross-sectional study stems from a comprehensive evaluation of the 2005 Quebec MH reform [1], which set up 95 local health services networks, and merged general hospitals, nursing homes, and local community service centers [all public services] to create a health and social service center [HSSC] within each network. The HSSCs handle multiple functions including the supervision of care quality within primary MH care services, as well as the coordination of primary care and specialized MH professionals within the respective networks.

The sample consisted of MH professionals selected from four local health service networks. The networks were chosen to include diverse geographical areas [urban or semi-urban], demographic characteristics [populations ranging from 135,000 to 300,000], and diversity of services offered. Two networks were located in a large metropolitan area inhabited by half the Quebec population, with specialized services available from a MH university institute. The third network was located in the provincial capital region, where another MH university Institute is located. The fourth network was located in a remote, semi-urban area, where a general hospital provided MH services. All the networks offered primary MH care, as well.

An advisory committee, including representatives from the four networks, provided the researchers with a list of MH team managers for each network, who, in turn, furnished lists of all MH professionals from their respective teams. To be eligible for the study, professionals had to be members of a MH team that included three or more professionals representing at least two disciplines, based on previous research related to teamwork, and taking into account a minimum of interaction patterns required by the complexity of MH care [47, 48]. MH professionals were approached by the research team via email, or telephone, regarding their potential interest and involvement in the research.

A self-administered questionnaire was mailed to 466 eligible MH professionals. Questionnaire items included 13 standardized scales on multidimensional aspects of teamwork, and nine questions on individual characteristics [e.g. age, gender]. Data collection occurred between May and November 2013, following three recruitment drives in the four networks. Team managers who employed professionals recruited to the study [n = 49, October 2013 and June 2014], were themselves recruited by email or telephone during the same period. Managers received a second questionnaire, which included information available in administrative databases, regarding eight elements: 1] manager characteristics [e.g. age, gender]; 2] client profiles [e.g. proportion of heavy service users]; 3] team attributes [e.g. size, setting]; 4] clinical activities [e.g. clinical approaches]; 5] organizational culture; 6] network integration strategies [e.g. service agreements]; 7] frequency and satisfaction of interactions with other network teams and organizations; and 8] MH services in the network. Only data on client profiles, team attributes and organizational culture were relevant to this study. Both manager and MH professional questionnaires were pre-tested by eight professionals, two from each network. All study participants signed a consent form. The Douglas Mental Health University Institute research ethics committee approved the multi-site research protocol [MP-IUMD-11037].

Conceptual framework, variables and standardized scales

The conceptual framework used in this study was adapted from the IMOI Model [22] and the ITEM [23], with independent variables organized within the five following domains: 1] Provider Characteristics, 2] Team Attributes, 3] Team Processes, 4] Team Emergent States, and 5] Organizational Culture [Fig 1]. Data regarding Professional Characteristics, Team Processes, and Team Emergent States [Fig 1] were mainly drawn from the MH professional questionnaire, while data on Team Attributes and Organizational Culture came from the manager questionnaire.

Conceptual framework.

Independent Variables were grouped into five domains: 1] Professional Characteristics; 2] Team Attributes; 3]; Team Processes; 4] Team Emergent States; and 5] Organizational Culture.

The dependent variable “Job satisfaction” was assessed using the French-language version of the Job Satisfaction Survey [49]. The original instrument consisted of 36 items grouped into nine sub-dimensions with Likert scale responses [Cronbach alpha = 0.91 for the global scale, between 60–78 according the sub-dimension [49]]. We eliminated 16 items and 4 sub-dimensions that dealt with remuneration, as this issue did not apply to the Quebec public healthcare system. Measures of Cronbach’s alpha for the scale used in this study ranged from 0.63 [for relations with co-workers] to 0.78 [for job conditions].

Independent variables on Professional Characteristics were: Age, Gender, Type of profession, Years of professional practice, and Seniority on the team. Types of profession included four groups: 1] Medical [psychiatrist, pharmacist, general practitioner], 2] Nurse, 3] Psychosocial [social worker, psychologist], and 4] General [technician, clerk].

Independent variables for Team Attributes were: Team composition [number, types of professionals], Team setting [Specialized outpatient MH teams, Specialized inpatient MH teams, Primary care teams] and Client profile [proportions with: severe MH disorders, personality disorders, co-morbid MH and substance use disorders, co-morbid MH disorders and chronic physical illnesses, suicide ideation, and heavy service users].

Independent variables for Team Processes were measured using eight standardized scales that assessed: Team support, Team conflict, Cross-functional integration [e.g. integration among various disciplines], Knowledge production, Informational role self-efficacy [i.e. belief in the ability to complete a task or deal with exterior demands], Participation in the decision-making process, Team autonomy, and Team collaboration. These scales were translated into French, and validated, with the exception of the scales for Informational role self-efficacy and Team collaboration that were originally developed in French.

Independent variables for Team Emergent States were measured using four standardized scales that assessed Trust, Affective commitment toward the team, Team climate, and Belief in the advantages of interdisciplinary collaboration. The relevant scales were translated into French, and validated, except for the scale on Belief in the advantages of interdisciplinary collaboration, which originally appeared in French.

Finally, Organizational Culture, as measured by team managers, was measured using the Organizational Culture Assessment Instrument [41], which consists of six questions. Reponses include four possible choices, with each response distributed among a possible 100 points. Organizational Cultures were classified as 1] Family/clan culture; 2] Entrepreneurial culture; 3] Market/rational culture; and 4] Hierarchical culture. Table 1 presents details of the standardized scales used in the study and their Cronbach’s alpha scores at original validation, and in the present study.

Table 1

Description of standardized instruments included in the study.

Analyses

Statistical analyses were run with SPSS, 24th edition. Missing values accounted for less than 5% per variable, and were randomly distributed. They were treated using Expectation Maximization, a multiple imputation technique. Following preparation of the database, we conducted univariate, bivariate and multivariate analyses. Univariate analyses included central tendency measures [mean value] for continuous variables and frequency distributions [numbers and percentages] for categorical variables. Normality assumptions were assessed for the dependent variable. Bivariate analyses were then performed using simple linear regression, and ANOVA t-tests, with the alpha value set at 0.10. Variables significantly associated with the dependent variable were used to build the multiple linear regression model, with alpha set at 0.05. The variance explained [R2] by the model, and goodness of fit [F test and p value], were also calculated.

To assess the effects of MH networks and MH teams on Job satisfaction, we ran a multilevel analysis using independent variables significantly associated with the dependent variable [Job Satisfaction] in the multiple linear regression model. An unconditional model was first run, using only the dependent variable. Intra-class correlation was then calculated, estimating the proportion of total variance in Job satisfaction attributable to the networks and to teams. Fixed and random coefficients were then tested, step by step, using maximum likelihood estimation. The fit was assessed for each model using maximum likelihood [-2LL], Akaike's information criterion [AIC] and Schwarz's Bayesian criterion [BIC]. The different models were compared one to another, with a smaller value indicating improvement in model fit. Covariance parameters were estimated using the Wald Z test, which assesses whether variability in the intercepts was significantly different among MH teams and the networks. As the Wald Z test is considered unreliable [60], results were not viewed as conclusive even when testing produced a non-significant P value. Finally, a multilevel model was generated and compared to the multiple linear regression model in order to highlight differences and decide whether the networks or teams contributed to a better assessment of Job satisfaction among MH professionals.

Results

Of 466 MH professionals recruited to the study, 315 participated, for a 68% response rate. The 315 MH professionals came from 49 MH teams, with six members on average [range from three to 16]. In all, the teams represented nine different health care groups. Specialized service groups included hospital units, day hospitals, assertive community treatment programs, outpatient clinics, and rehabilitation programs. Primary care groups included evaluation units, local community service centers, basic teams and intensive case management programs. An average of 35 professionals [ranging from 30 to 55] worked in each group. Response rates for the networks were 64% [157/247] for networks 1 and 2; 59% [117/198] for network 3; and 80% [44/55] for network 4.

There were no differences between participant and non-participant MH professionals on the distributions for team type [χ2 [1, N = 466] = 0.79; p = 0.68], or gender [χ2 [1, N = 466] = 0.03; p = 0.87]. The mean age was 43, and mean seniority on clinical teams was three years. Women outnumbered men by more than two to one [70% versus 30%]. Most participants [78%] worked full-time, while 22% worked part-time. More than half [55%] were Psychosocial professionals, followed by Nurses [30%], General workers [11%] and Physicians [4%]. Most worked in Specialized outpatient MH teams [56%], with the remainder in either Primary care teams [32%] or Specialized inpatient MH teams [12%]. Almost three-quarters of the teams [74%] operated in the three urban settings, versus 16% in the semi-urban setting. The outcome variable “Job satisfaction” had a mean score of 24.8 [range: 11.3–35.0; SD: 3.6] and was normally distributed [skewness:–.037; kurtosis: .332].

Regarding team managers, 41 participated out of 49 invited to the study, for a response rate of 84%. Response rates within the networks were 75% for networks 1 and 2 [18/24]; 94% for network 3 [16/17]; and 88% for network 4 [7/8]. Comparative analyses revealed no differences between respondent and non-respondent managers on gender [Pearson chi-square = .966; df = 1; Fisher’s exact test two-sided p = .663]; or type of health care setting [Pearson chi-square = 1.861; df = 1; Fisher’s exact test two-sided p = .245.] Of participating managers, 71% were female, 62% were members of Specialized MH teams, and 38% worked in Primary care teams. The mean age was 44, and mean seniority in the team four years.

Table 2 presents 15 variables linked to job satisfaction in the bivariate analysis. Four of these were positively related to Team Processes: Team support, Cross-functional integration, Knowledge production, Informational role self-efficacy; whereas one variable, Team conflict, was negatively related. Four variables in Table 2 were related to Team Emergent States: Trust, Affective commitment toward the team, Team climate, and Belief in advantages of interdisciplinary collaboration. Three variables emerged under Team Attributes: Proportion of personality disorder in the clientele, Specialized inpatient MH teams and, marginally, Proportion of co-morbid MH disorder and chronic physical illnesses in the clientele. Job satisfaction was associated with two types of Organizational Culture, as measured by team managers: Market/rational culture and, negatively, with Family/clan culture. Finally, only one variable among Professional Characteristics [Male gender] was related to Job satisfaction.

Table 2

Participant characteristics and unadjusted associations with job satisfaction [N = 315.

Table 3 presents the multiple linear regression model for the study. Four variables were identified in the model as independently and positively associated with Job satisfaction: Team support, Team collaboration, Participation in the decision-making process [Team Processes], and Affective commitment toward the team [Team Emergent States]. Two variables were negatively associated with Job satisfaction: Team conflict [Team Processes] and Market/rational culture [Organizational Culture]. No variables under Professional Characteristics or Team Attributes were linked to Job satisfaction in the final model. This model explained 45% of the total variance and had acceptable goodness of fit.

Table 3

Variables independently associated with job satisfaction: Multiple linear regression.