We used a cross-sectional survey method and selected five provinces (Heilongjiang, Sichuan, Anhui, Gansu, and Shandong) as survey sites to recruit subjects from May to July 2022. Using nonrandom convenience sampling, we recruited subjects from the community who met the survey criteria. The selected study subjects filled out an anonymous questionnaire via an online platform (Survey Star, Changsha Ran Xing Science and Technology, Shanghai, China). The key variables in the questionnaire were all required and assigned logical values. Data were screened according to the requirements of the study, finally the questionnaire information of 1,561 cases were selected, and then the database was locked.

The inclusion criteria for survey respondents included the following: they needed to be PHCWs who had online access so they could complete the survey. Participation was voluntary.

Sample size calculation was based on the cross-sectional survey design. The overall burnout indicator for healthcare work obtained from the data was approximately p = 0.3, α = 0.05, and d = 0.1 × p = 0.03. The sample size for a purely random sample was derived from the formula for cross-sectional survey sample size. Considering the sample size expansion (1.5–2.0 times) problem for nonrandom sampling, the minimum sample size is expanded to N_srs = 897 × 1.5 = 1,346:

The questionnaire was divided into four parts:

Burnout was measured using the MBI scale, quantified using the Likert-type scale, and evaluated according to the SS′ scoring principle: SS′ = 0.4 × mean score for emotional exhaustion + 0.3 × mean score for depersonalization + 0.3 × (6 – mean score for personal accomplishment) (26, 27). Based on the scores, the subjects were divided into three categories: (1) no job burnout (0 ≤ SS′ < 1.50), (2) mild job burnout (1.50 ≤ SS′ < 3.50), and (3) severe job burnout (3.50 ≤ SS′ < 6).

In this study, mild and severe job burnout are regarded as the levels of job burnout that are in need of improvement; that is, the detection rate of job burnout is positive:

The questionnaire was analyzed using R 4.1.2 (R Development Core Team) and IBM SPSS AMOS 26.0.0 (IBM Corporation, Armonk, NY, USA). Differences were statistically significant at p < 0.05.

We used the chi-square test to analyze the correlation between the job burnout level of healthcare workers and demographic factors. Stepwise logistic regression was used to analyze the factors affecting job burnout. We established a structural equation model (SEM) based on theoretical assumptions and the factors affecting burnout to explore the path coefficients of potential variables influencing burnout. The generalized least-squares (GLS) method was used to estimate the path coefficients. We computed the fit of the model to the data using the following: chi-squared/degree of freedom (CMIN/df), root-mean-square error of approximation (RMSEA), goodness-of-fit index (GFI), adjusted goodness-of-fit index (AGFI), and comparative fit index (CFI). Furthermore, multigroup SEM was used to explore similarities and differences in the model according to age, gender, years of work, and whether engaged in new COVID-19-related work.

We used Thurstone's pairwise comparison method to analyze the ranked items of burnout improvement methods. In this method, option combination information is converted into pairwise comparison information, and the value of column j for row i is divided into three cases: R_α, R_β, and R_γ. The formula for calculating the probability table p_ij and the scale value S_i is

The overall Cronbach's α coefficient of the Chinese version of the MBI scale in this study was 0.859. The internal Cronbach's α coefficients of the dimensions of emotional exhaustion, depersonalization, and personal accomplishment were 0.926, 0.914, and 0.843, respectively. The split-half reliabilities of emotional exhaustion, depersonalization, and personal accomplishment were 0.869, 0.712, and 0.878, respectively.

The χ² value of Bartlett's sphericity test was 20,333.90, p < 0.001. The Kaiser–Meyer–Olkin measure of sampling adequacy was 0.922. Three factors with the characteristic root λ > 1 were extracted by exploratory factor analysis, and the cumulative variance contribution rate was 73.01%. Factors 1, 2, and 3 explained the three dimensions of personal accomplishment, depersonalization, and emotional exhaustion in the MBI scale, respectively (Supplementary Table S1).

The option ranking method was used for the preferred ways to reduce job burnout. We presented five ways to reduce job burnout: (1) awarding an honorary certificate, (2) reducing work intensity, (3) improving the work environment, (4) providing opportunities for further study, and (5) Increasing wages.

Increasing wages (88.68%) and improving the work environment (83.09%) were found to be most effective. Analyzing the combination of options, most healthcare workers (34.98%) reduced their work intensity, improved their work environment, and increased their salaries. According to the ranking analysis of the importance of the options, the most effective healthcare workers (64.95%) can improve the work environment (Figure 4).

Based on the Thurston method, the scale values of the five options (1, 2, 3, 4, and 5) were ranked on a psychological valence chart. The results showed that the most-preferred ways for PHCWs to reduce job burnout were improving the work environment (0.913) and increasing salaries (0.810). Meanwhile, the scale values of awarding honorary certificates or titles (−0.342), reducing work intensity (−0.594), and providing opportunities for further study (−0.787) were all negative (Supplementary Figure S1).

The MBI scale has been widely used to measure job burnout. During COVID-19, this scale was used in Italy, the US, Belgium, India, Singapore, and other countries to measure the job burnout of frontline healthcare workers, and its reliability and validity were verified (29–32).

We used a modified Chinese version of the MBI burnout scale to conduct a presurvey and test scale reliability and validity at 10 sites in Heilongjiang, Sichuan, Gansu, Anhui, and Shandong Provinces. The Cronbach's α coefficients for emotional exhaustion, depersonalization, and personal accomplishment were >0.7, demonstrating that the scale had good reliability, internal consistency, and external consistency.

The overall reported job burnout rate among PHCWs was about 60%; mild burnout accounted for 55%, and severe burnout accounted for 5%. Galanis summarized 16 studies where the MBI was used to measure nurses' job burnout and found that the emotional exhaustion rate was 34%, the personality disintegration rate was 12%, and the low personal achievement rate was 15% (33). Compared with previous findings, the job burnout rate among PHCWs in China was found to be significantly higher (34, 35), suggesting that this issue warrants attention in China.

The results suggested that job burnout among PHCWs is characterized by high emotional exhaustion, high personality disintegration, and low personal accomplishment, among which the score for personal accomplishment was significantly below the norm. This is similar to the findings of Hu et al. (36) and Parandeh et al. (37). It can be attributed to the fact that PHCWs in China were mostly engaged in repetitive tasks with high work pressure and long work hours during the pandemic (38) and may have encountered unsupportive or uncooperative patients (39).

PHCWs played an important role in the struggle to contain COVID-19 (40). From 2020 to 2022, China's approach to the pandemic shifted from “zero clearing” to “dynamic clearing” (41, 42). PHCWs took on the tedious work of COVID-19 patient treatment, nucleic acid testing, epidemiological investigation, vaccination, isolation and prevention, and decontamination (43, 44), which involved long work hours and high work pressure.

The factors affecting burnout were diverse and changed over time, and could be both subjective and objective. Our findings showed that burnout level was associated with the age, years of work, education level, and work location of healthcare workers. Among them, age ≥ 35 years, better work environments, and more professional pride inhibited emotional exhaustion while more intense work exacerbated it. Gambaro et al.'s study on job burnout of healthcare workers also showed a negative correlation between age and job burnout. Similar to our findings, work experience has been shown to supply healthcare workers with the knowledge and emotion regulation skills they need to mitigate burnout (45). During the COVID-19 pandemic, healthcare staff with higher titles often had to assume more responsibilities, thus leading to emotional exhaustion (8, 46).

The depersonalization dimension was similar to that of emotional exhaustion. Therefore, PHCWs who are younger, have higher technical titles, and have higher work intensity should be the focus of interventions for burnout. In addition, females were less likely to show symptoms of depersonalization (i.e., holding negative or inappropriate attitudes toward their work objects) (47). Previous studies have also shown that female healthcare workers have more empathy for patients, better understand patients, and devote more time to them (48–50).

In the personal accomplishment dimension, healthcare workers with more than 10 years of work experience had lower levels of personal fulfillment. Studies have shown that longer years of work are usually a contributing factor to burnout (51, 52). Different from previous studies, we found that those with high work intensity showed a higher level of personal accomplishment (36). This reflects the sense of social responsibility and dedication shown by Chinese healthcare workers during the pandemic (38).

Our results highlight the important role of job satisfaction in reducing healthcare workers' job burnout. SEM showed that improving job satisfaction could reduce job burnout. Our findings partly confirm Goulet's and Singh theory of career commitment—that is, job satisfaction has a negative effect on job burnout (53). Therefore, as an important factor affecting PHCWs' professional development, job satisfaction should be an important intervention strategy in occupational health (54). Social status also affects the job satisfaction of healthcare workers, thus affecting job burnout. Thus, more attention should be paid to groups with higher social status (38).

Among the ways to reduce the job burnout among PHCWs, improving the work environment and increasing wages are the most important. During the COVID-19 pandemic, healthcare workers often worked in isolation wards and temporary nucleic acid test sites, and the work environments were relatively harsh, which could easily lead to burnout (38). Healthcare workers expressed the most dissatisfaction with the remuneration dimension. PHCWs in China have low salaries but bear higher workloads, greater risks of infection, and heavier physical and mental pressure (8, 55, 56). Incentive policies should be implemented to improve healthcare workers' job satisfaction and alleviate burnout by increasing their income.