Prevalence and Risk Factors of Burnout Syndrome during COVID-19 Pandemic among Healthcare Providers in Thailand

Objective To study prevalence, risk factors, and consequences of the COVID-19 pandemic related to Burnout syndrome (BOS) among Thai healthcare providers (HCPs) during the COVID-19 pandemic. Methods We performed a cross-sectional study among HCPs, involved in caring for patients during the pandemic in two periods (1st period, May–Jun 2021, and 2nd period, Sep-Oct 2021). Data were distributed using electronic questionnaires. BOS was defined if respondents exhibited a high level of at least one domain in the Maslach Burnout Inventory criteria. The primary outcome was prevalence of BOS. Results Altogether, 2,027 and 1,146 respondents were enrolled in the 1st and 2nd periods, respectively. Most respondents were female (73.3, 68.2%). The top three job positions were physicians (49.2, 58.9%), nurses (41.2, 30.6%), and nursing assistants (4.8, 6.5%), respectively. No difference was found in overall prevalence of Burnout syndrome during the 1st and 2nd periods (73 vs. 73.5%, p=0.80). Using multivariate analysis, significant risk factors for Burnout syndrome in both periods were (1) living with family (odds ratio (OR) 1.3 and 1.5), (2) tertiary care hospital (OR 1.92 and 2.13), (3) nurse (OR 1.38 and 2.29), (4) nursing assistant (OR 0.92 and 4.81), (5) salary ≤40,000 THB (OR 1.53 and 1.53), (6) >20 patients per shift (OR 1.55 and 1.88), (7) >6 shifts after hours monthly (OR 1.26 and 1.49), and (8) ≤1 rest day weekly (OR 1.3 and 1.4). Conclusion We found a high prevalence of Burnout syndrome among Thai HCPs during the pandemic. Knowing those risk factors may provide a strategy to BOS during the pandemic.


Introduction
Burnout syndrome (BOS) is a work-related stress syndrome described as diminished interest and exhaustion caused by experiencing an inability to cope with emotional stress at work [1,2]. Tis syndrome has been frequently associated with specifc consequences for individuals (depression, lack of motivation, and workplace violence in a situation mutual with burnout) and also for health institutions (shortage of healthcare providers (HCPs), strain of healthcare systems, malpractice litigation, and poor quality of care) resulting in decreased organization efciency [3][4][5][6][7][8][9]. BOS has been recognized as a psychological problem and has become much more prevalent in the last decade; it was defned as consisting of three qualitative dimensions, namely, emotional exhaustion (EE), depersonalization (DP) or cynicism, and reduced personal accomplishment (PA) [10][11][12][13]. BOS develops slowly and could be triggered by multiple factors. However, it has almost never been identifed in its early stages [10,14].
Before the emergence of coronavirus disease 2019 (COVID- 19), the prevalence of BOS among HCPs was found to be between 25 and 62.6% and varied across countries [15][16][17][18][19][20][21][22][23][24][25][26][27]. Currently, the global pandemic of COVID-19 has caused numerous infected cases and deaths [28], a rapidly increasing number of patients, high workload, shortage of medical resources, the pain of losing patients and colleagues, and the risk of infection for themselves and their families may cause more BOS among HCPs, but the efect of the pandemic on BOS is yet poorly understood. Terefore, this study aimed to evaluate the prevalence and risk factors of BOS among HCPs during the COVID-19 pandemic in Tailand.

Methods and Materials
2.1. Participants. We conducted a cross-sectional analytic study among HCPs (physicians, nurses, nursing assistants, pharmacists, laboratory staf, and radiology staf) during the COVID-19 pandemic in Tailand. Based on social distancing policy during the COVID-19 pandemic, participants were recruited through online social media platforms (Google Forms) using a snowball technique. Te inclusion criteria included full-time employees at all types of medical centers in Tailand, including university hospitals, tertiary center hospitals, secondary center hospitals, community hospitals, private hospitals, and private clinics, who were willing to answer a questionnaire. Either general medical centers or specialized COVID-19 centers were included in this study. Exclusion criteria included (1) [29]. Terefore, to compare and analyze whether more COVID-19 hospitalized cases increased the prevalence of BOS, we decided to continue collecting data from September to October 2021. We defned the data during May 1 to June 30, 2021, as the 1st period while data from September 1 to October 31, 2021, were classifed as the 2nd period. We also tested whether risk factors for BOS during the frst period were also replicated during the second period.

Questionnaire.
Tis study was conducted using a Tai language electronic survey-based questionnaire consisting of fve parts: (i) demographic data, (ii) personal information, (iii) work characteristics, (iv) consequences related to COVID-19, and (v) Maslach Burnout Inventory (MBI) Human Service survey. Tus, we collected gender, age, current marital status (single, married, and divorce), and accommodation status (single living, living with family, and living with colleague/friend). In terms of the work characteristics, participants were required to verify their type of workplace (university hospital, tertiary center hospital, secondary center hospital, community hospital, private hospital, private clinic, or others), job position (physician, nurse, nursing assistant, pharmacist, laboratory staf, radiologic staf, administrative staf, or others), type of specialist, work experience duration, range of monthly income, and current workload including the number of patients, work hours, and days of. Regarding the consequences of COVID-19 pandemic related to BOS, the questionnaires comprised workload, position change, concerns of family health, and exhaustion from using personal protective equipment. Te consequences of COVID-19 pandemic related to BOS were verifed if the participants responded "yes" to the following questions: "Is your workload heavier?" "Have you had to change your job to less satisfactory one?" "Do you have any concerns about your health?" "Do you have any concerns about your family's health?" "Do you fnd it difcult to live in your daily life in the COVID-19 pandemic?" "Are you exhausted from wearing personal protective equipment (PPE) whilst caring COVID-19 patients?".
As it is considered an internationally acknowledged and validated instrument for measuring job burnout, MBI Human Services Survey is a 22-item questionnaire, in which the answers are self-graded frequency score from 0 (never) to 6 (every day). All 22 questions are categorized in 3 dimensions, including EE, DP, and PA. Te summation for each aspect is stratifed as high, average, or low. BOS is defned if the participants revealed high EE, high DP, or low PA [30][31][32][33][34]. Te MBI Human Service survey has previously been translated into Tai and has been well tested for reliability with Cronbach's alpha coefcients for EE of 0.92, DP of 0.66, and PA of 0.65 [35,36].

Statistical Analysis.
According to the related study of BOS during the COVID-19 pandemic conducted by Nishimura et al. [37], adequate sample size in our study must comprise at least 385 participants.. Primary outcome was prevalence of the BOS in HCPs during the COVID-9 pandemic. Secondary outcomes consisted of risk factors for the BOS and consequences of the COVID-19 pandemic related to BOS. Descriptive statistics, including percentage, frequency, average, and standard deviation, were used to demonstrate the demographic data and prevalence of BOS. To analyze risk factors for BOS, we performed univariate comparison frst then constructed multivariate regression analysis. Inferential statistics were used by exhibiting a statistical signifcance at alpha of 0.05, including Pearson's chisquare test and independent-student t-test.

Prevalence of BOS.
Although the overall prevalence of BOS was high in both study periods, no diference was found in overall prevalence of BOS between the 1st (73%) and the 2nd period (73.5%) (p � 0.8). In the 1st and 2nd periods,  Figure 1). In the subgroup analysis based on physician specialty, no diference was found in the prevalence of BOS between the two study periods, except among other specialties ( Figure 2 (Table 3).

Consequences of COVID-19 Pandemic Related to BOS.
Using multivariate analysis, consequences of COVID-19 pandemic related to BOS were heavier workload, unsatisfed job, personal and family health concerns, difculty in daily life, and exhaustion of wearing personal protective equipment (all p < 0.05 in both periods) ( Table 4).

Discussion
Tis study demonstrated a high prevalence of BOS in HCPs during the COVID-19 pandemic in Tailand. We found a high prevalence of BOS in both periods of the study (73.0 and 73.5%). Compared to a previous study before the COVID-19 pandemic, the prevalence of BOS among Tai ICU physicians and nurses was 65.15 and 60.95%, respectively [23] while a previous systematic review during the COVID-19 pandemic, the pooled overall BOS prevalence was 52% (95% CI 40-63%) [38]. We also found that the major domain contributing to BOS was high EE in the 1st period and high level of decreased PA in the 2nd period. Tese fndings were aligned to a previous study which found that most of Italian HCPs experienced high EE and decreased PA during the COVID-19 pandemic [39]. Physicians in our study experienced lower BOS in the 2nd period but varied between specialties. Although it did not achieve statistical signifcance, pulmonologists and pediatricians seem increased prevalence of BOS in the 2nd period. Te possible explanations may be most severe COVID-19 cases were taken care by pulmonologists. Increasing numbers of patients have may put more severe cases as well as higher workload on the pulmonologists. Likewise, a greater number of children with COVID-19 infection in the 2nd period may have contributed to higher pediatrician workloads. Te nurses and nursing assistants dealing with a higher prevalence of BOS in the 2nd period may have  Physician specialties Figure 2: Prevalence of Burnout syndrome classifed by physician specialties. White and black bar graphs represent prevalence of Burnout syndrome (BOS) in the 1st and 2nd periods, respectively. Numbers inside the bar graphs represent the total number of respondents in each physician's specialty. Two dash lines represent average BOS in the 1st (71%) and 2nd periods (67%), respectively. * depicts p < 0.05, data analysis using the Pearson chi-square test. CCM: critical care medicine; Pulmo: pulmonologist; ID: infectious disease physician; EM: emergency medicine; Anesth: anesthesiologist; Med: internal medicine; Sx: general surgery; Eye: ophthalmologist; ENT: otolaryngologist; Ped: pediatrician; a Others: physical medicine and rehabilitation physician/obstetrician/radiologist/family medicine physician/pathologist/ psychiatrist.  Journal of Environmental and Public Health possibly been due to more hospitalized COVID-19 cases in the 2nd period of the study and the increased burden of nursing care. Te risk factors for BOS we found in our study included (1) living with family, (2) working in a tertiary care hospital, (3) nurse position, (4) nursing assistant position, (5) earning salary ≤40,000 THB, (6) having >20 patients per shift, (7) having >6 shifts after hours monthly, and (8) having ≤1 rest day weekly. Tese important fndings provide public health strategies to prevent BOS in HCPs during the COVID-19 pandemic. Even though workload is an uncontrollable factor during the pandemic, organizations need to ensure adequate staging through ongoing evaluation of workloads. Optimization of professional rewarding systems, including salaries and motivational support such as temporary accommodation or family welfare, should also be considered. Although the role of family relations may be less pronounced than job satisfaction itself, some studies found that families can take some action in the mitigation of occupational BOS in HCPs [26,40,41]. In contrast, our study found that living with family during the pandemic increased the risk of BOS. Tis fnding is possibly explained by the concern of family members becoming infected through HCPs. A previous study in Indian HCPs during the COVID-19 pandemic [42] used a questionnaire to evaluate BOS. Tey found that 50% of participants had a statistically signifcant concern and reported exhaustion from working during the pandemic. Tose authors also worried about becoming infected and transmitting the disease to their families. One more risk factor for BOS in our study was working in a tertiary care hospital; a possible explanation why working in the tertiary care hospitals increased risk of BOS may be because of receiving very severe cases from other hospitals which have less facility. Also, inappropriate ratios of HCPs to number of patients as well as insufcient medical equipment compared to the number of patients may play a major role to BOS. Other risk factors such as low salary, increased workload, and less rest days weekly also increased risk for BOS because these factors might reduce motivation to work.

Journal of Environmental and Public Health
Compared to previous studies, this study revealed novel fndings. We demonstrated that the COVID-19 pandemic is directly linked to BOS through both physical and mental drawbacks. For example, we found that HCPs with BOS were signifcantly exhausted from wearing PPE whilst caring for COVID-19 patients, and had more concerns about their and their family's health. Providing several specifc preventive interventions during a pandemic may be a key to potential BOS reduction, including measures such as confrmation of correct PPE usage, emphasizing ways to reduce the risk of infection, supplying adequate antigen test kits for early detection, and providing isolated accommodation whilst on duty.
Te strengths of our study were a large number of participants with various job positions and various hospital levels in Tailand. Tis allows our data to be more diverse. We also validated possible risk factors for BOS and consequences of COVID-19 pandemic related to BOS using the 2nd period cohort as a validation cohort with the multivariate analysis. However, our study encountered several limitations. Firstly, respondents in both periods of the study difered which may have afected the comparison of results. Secondly, response bias may have occurred comprising their personal data. Lastly, our data must be carefully interpreted because of heterogeneity in job positions, types of hospital, and various diferences in baseline characteristics (i.e., marital status and monthly income).
Te fnding of high prevalence of BOS in our study may refect overall growing disappointment in healthcare systems, which need urgent attention because it may result in mental health problems, decreased quality of life among HCPs, and poorer healthcare outcomes [43,44]. Moreover, preventive interventions, including occupational health surveillance and workplace health promotion programs, should be reviewed for prevention, early diagnosis and therapy of BOS, as well as other pandemic stress-related consequences such as posttraumatic stress disorder and suicide [45][46][47][48]. Te issue of long-term BOS and mental health problems among HCPs should be examined further and big data intelligence in COVID-19 pandemic may play an important role for further research [49].

Conclusion
We found a high prevalence of BOS among Tai HCPs during the COVID-19 pandemic. Tis needs more attention, preventive intervention, and increased public health awareness to reduce BOS in HCPs. Increased workload with low compensation is a major risk factor for BOS. In addition, consequences of COVID-19 pandemic signifcantly related to BOS among HCPs.

Data Availability
Te data that support the fndings of this study are available on request from the corresponding author. Te data are not publicly available because of privacy or ethical restrictions.

Ethical Approval
Tis study was approved by the Ethics Committee Institutional Review Board of Royal Tai Army Medical Department (R075q/64_Exp).

Consent
Te informed consent was obtained from all participants.

Conflicts of Interest
All authors report no fnancial or other relationships that represent actual or potential conficts of interest relevant to the content of this study.