Study design and study population
The Chinese Sixth National Health and Services Survey (NHSS) in Sichuan Province was conducted from August 2018 to October 2018. All healthcare workers who had a practicing qualified certificate on file in third-class comprehensive hospitals, second-class hospitals, community health service centres, and township hospitals of Sichuan Province were eligible to be study subjects. Multistage stratified random sampling was used to acquire the study sample. In the first stage, 14 cities were randomly selected from among 21 prefecture-level cities, and 70 towns and streets were randomly selected from the 14 cities. In the second stage, all third-class comprehensive hospitals and some of the second-class hospitals were selected from the 14 cities, and all the community health service centres and township hospitals in the 70 towns and streets were enrolled in this survey. In the third stage, 20 physicians and 10 nurses were selected from each third-class comprehensive and second-class hospital by a simple random sampling method. Five physicians, three nurses, and two public health professionals were randomly selected from each community health service centre and township hospital. Respondents were asked to fill out the questionnaire anonymously. Informed consent was obtained from each healthcare worker following a detailed explanation about the purpose of the study. Overall, 1685 healthcare workers were investigated; we obtained and analysed 1327 (78.80%) valid responses.
Measures
The questionnaire was designed by an expert panel from the National Health Commission of the People’s Republic of China.
Socio-demographic characteristics of healthcare workers
The socio-demographic characteristics examined included: gender, age (<30, 30–39, 40–49, and ≥ 50 years), marital status (single, divorced, widowed, or married), education level (junior college or below, bachelor, master, or above), specialty (physician, nurse, and public health professional), technical title (no title, primary title, middle title, vice-senior title,or above), working years (<5, 5–9, 10–19, 20–29, and ≥30 years), weekly hours at work (≤40h, >40h), night shifts per month (none, 1–7, and ≥8), grade of medical institutions (community health service centres and township hospitals, second-class hospitals, and above).
Effort-reward imbalance
The ERI scale includes three dimensions: extrinsic effort (3 items), reward (7 items) and overcommitment (6 items). The responses to all the items were scored on a four-point Likert scale (1=strongly disagree, 4=strongly agree). To assess the degree of imbalance between high cost and low gain at work, effort-reward ratio (ERR) was calculated as: E/(R*C). E was the total score of the effort dimension, R was the total score of the reward dimension, and C was the correction coefficient based on the difference in the number of numerators and denominators [27]. Here, C=3/7=0.4286. An ERR value of >1.0 indicates that the amount of effort is not rewarded adequately [19]. Higher scores represented higher overcommitment to work. The Cronbach’s alpha coefficient of this scale in this study was 0.786.
Job Satisfaction
Work Engagement
Work engagement was measured by the Chinese version of the Utrecht Work Engagement Scale [28]. It comprised seventeen items measuring three aspects of work engagement: work vigour (6 items), work dedication (5 items), and work absorption (6 items). Responses were given on a seven-point Likert scale ranging from 0 (never) to 6 (every day) and were combined into summary scores. Higher scores indicated higher work engagement. The Cronbach’s alpha coefficient of this scale in this study was 0.941.
Outcome variable
Self-rated health status was assigned scores of 5(good), 4(relatively good), 3(fair), 2(relatively poor), and 1(poor) by asking the participants ‘How do you feel about your health?’ Higher scores indicated better self-rated health.
Statistical Analysis
We first used descriptive statistics to examine the socio-demographic characteristics, ERR, overcommitment, job satisfaction, work engagement, and self-rated health among healthcare workers. Second, Pearson’s correlation coefficient was used to analyse the correlations among ERR, overcommitment, work engagement, job satisfaction, and self-rated health. Third, a structural equation model (SEM) was employed to further test the hypothesized relationships among effort-reward imbalance, job satisfaction, work engagement, and self-rated health among healthcare workers.
Several indicators were used to assess the fit between the current data and the hypothesized model, namely goodness of fit index (GFI) >0.9, norm fit index (NFI) >0.9, relative fit index (RFI) >0.9, comparative fit index (CFI) >0.9, incremental fit index (IFI) >0.9, Tucker-Lewis Index (TLI) >0.9, which indicated that the model fit was acceptable. All statistical analyses were performed using the IBM SPSS version 23.0 (SPSS Inc., Chicago, IL, USA) and Analysis of Moment Structures (AMOS) version 22.0 (IBM, New York, NY, USA). The statistical significance was set at P <0.05.