If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Address correspondence to Benjamin Y.Q. Tan, MBBS, MRCP, National University Health System, NUHS Tower Block Level 10, 1E Kent Ridge Rd, Singapore 119228.
Yong Loo Lin School of Medicine, National University of Singapore, SingaporeDivision of Neurology, Department of Medicine, National University Hospital, Singapore
The strain on health care systems due to the COVID-19 pandemic has led to increased psychological distress among health care workers (HCWs). As this global crisis continues with little signs of abatement, we examine burnout and associated factors among HCWs.
Design
Cross-sectional survey study.
Setting and Participants
Doctors, nurses, allied health professionals, administrative, and support staff in 4 public hospitals and 1 primary care service in Singapore 3 months after COVID-19 was declared a global pandemic.
Methods
Study questionnaire captured demographic and workplace environment information and comprised 3 validated instruments, namely the Oldenburg Burnout Inventory (OLBI), Safety Attitudes Questionnaire (SAQ), and Hospital Anxiety and Depression Scale (HADS). Multivariate mixed model regression analyses were used to evaluate independent associations of mean OLBI-Disengagement and -Exhaustion scores. Further subgroup analysis was performed among redeployed HCWs.
Results
Among 11,286 invited HCWs, 3075 valid responses were received, giving an overall response rate of 27.2%. Mean OLBI scores were 2.38 and 2.50 for Disengagement and Exhaustion, respectively. Burnout thresholds in Disengagement and Exhaustion were met by 79.7% and 75.3% of respondents, respectively. On multivariate regression analysis, Chinese or Malay ethnicity, HADS anxiety or depression scores ≥8, shifts lasting ≥8 hours, and being redeployed were significantly associated with higher OLBI mean scores, whereas high SAQ scores were significantly associated with lower scores. Among redeployed HCWs, those redeployed to high-risk areas in a different facility (offsite) had lower burnout scores than those redeployed within their own work facility (onsite). A higher proportion of HCWs redeployed offsite assessed their training to be good or better compared with those redeployed onsite.
Conclusions and Implications
Every level of the health care workforce is susceptible to high levels of burnout during this pandemic. Modifiable workplace factors include adequate training, avoiding prolonged shifts ≥8 hours, and promoting safe working environments. Mitigating strategies should target every level of the health care workforce, including frontline and nonfrontline staff. Addressing and ameliorating burnout among HCWs should be a key priority for the sustainment of efforts to care for patients in the face of a prolonged pandemic.
The first COVID-19 case was reported in Singapore on January 23, 2020, and by April 3, 2020, stay-at-home orders, locally termed “circuit breaker” measures, were instituted whereby work-from-home arrangements were encouraged and schools and nonessential services such as dining, retail, and entertainment establishments were closed for approximately 2 months. Nonurgent medical care was reduced wherever possible to redeploy health care resource towards surge capacity and frontline units such as the emergency department, inpatient pandemic, and intensive care units.
Most semi-skilled workers involved in the construction, shipbuilding, and manufacturing industries are housed in dormitory-style accommodations regulated by the government,
which house between 3000 and 25,000 residents and feature communal living and various shared facilities, including laundry, recreation, eating, and groceries. In spite of our migrant workers comprising only 5% of Singapore's resident population, outbreaks among dormitory residents have contributed toward 95% of Singapore's total number of cases.
staffed by health care workers (HCWs) redeployed from public health care institutions and supplemented by private health care groups and volunteers. Other sites of HCW redeployment were Community Care and Swab Isolation facilities which were set up to isolate and care for clinically-well cases (potential- or proven-COID-19 positivity) who do not require inpatient care.
showed that when comparing high- versus low-risk exposure groups, the odds ratio for acute or posttraumatic stress (PTS) and psychological distress was 1.71 and 9.94, respectively. Similar findings were seen in the SARS outbreak of 2003
Psychological impact of the 2003 severe acute respiratory syndrome outbreak on health care workers in a medium size regional general hospital in Singapore.
when emotional distress, depression, and anxiety occurred more frequently among frontline HCWs. Burnout is a syndrome caused by chronic workplace stress and, according to Maslach and Jackson,
consists of 3 dimensions: emotional exhaustion, depersonalization, and feelings of reduced personal accomplishment. Burnout in HCWs has consistently shown a dose-response relationship with poorer patient safety outcomes.
Of note, burnout among HCWs during a pandemic is not fully understood, especially with regard to different facets of the work environment and concomitant psychological responses, such as anxiety and depression.
Pandemics require HCWs to sustain a period of heightened workload under stressful conditions, rapidly changing guidelines and redeployment to unfamiliar, high-risk settings
high per capita case load has necessitated a rapid redeployment of HCWs to FAST teams to staff medical posts under physically demanding conditions, such as under tentage in full personal protective equipment (PPE) amid Singapore's hot and humid climate with daytime temperatures reaching 35○C. At 6 months into this pandemic and no clear end in sight, we hypothesize that there would be a high level of burnout among HCWs that may be associated with changes in working environment including redeployment and workplace safety as well as anxiety and depression.
Methods
We conducted a multicenter, cross-sectional survey study whereby a questionnaire was distributed to doctors, nurses, allied health care professionals (AHPs), administrative or managerial staff, and support staff across 4 public hospitals involved in the care of COVID-19 cases (bed capacities of approximately 300, 700, 1200, and 2000 beds, respectively) and a public primary health care service from May 29 to June 24, 2020 (Supplementary Table 1). This was approximately 4 months from Singapore's first case and 2 months since the institution of national stay-at-home measures to curb COVID-19 spread. Support staff comprised nonclinical employees who were involved in the operations of the hospital such as porters, cleaners, laboratory technicians, and security staff. This questionnaire was voluntary, anonymous, worded in English, and distributed via corporate e-mail accounts using a secure, online questionnaire platform FormSG (GovTech, Singapore). As English is the standard operating language in Singapore, no literacy issues were encountered. The questionnaire captured basic demographic and workplace environment information and responses to 3 validated questionnaires: The Safety Attitudes Questionnaire (SAQ), Oldenburg Burnout Inventory (OLBI), and Hospital Anxiety and Depression Scale (HADS). We grouped respondents by categories such as (1) HCW roles, (2) Sex, (3) Ethnicity, (4) Redeployment outside primary roles, (5) Being tested for COVID-19, (6) Primary site of work, (7) Educational status, and (8) Average duration of shift during the pandemic (compared with routine 8- to 10-hour shifts during nonpandemic periods).
As redeployed HCWs constituted a large proportion of our health care workforce, we performed a subgroup analysis whereby this group was further divided into (1) Redeployed onsite (low risk), that is, not in direct contact with known COVID-19 cases; (2) Redeployed onsite (high risk), that is, highly likely to be in direct contact with known COVID-19 cases (eg, emergency department, critical care unit); and (3) Redeployed offsite to high-risk areas (eg, foreign work dormitory, community care facility, swab isolation facility).
Oldenburg Burnout Inventory
The OLBI is a 16-item validated tool to assess burnout
(Supplementary Table 2) covering 2 dimensions: Exhaustion and Disengagement. Disengagement refers to distancing oneself from the objects and content of one's work while exhaustion refers to feelings of emptiness, overwork, a strong need for rest, and physical exhaustion. Each dimension consists of 8 items rated on a 4-point Likert scale with options including “Strongly disagree,” “Disagree,” “Agree,” and “Strongly agree” with 4 points for the highest burnout response and 1 point for the lowest. The means and SDs were calculated for each domain and compared across baseline respondent characteristics. Burnout was determined with a cutoff of ≥2.25 for Exhaustion and ≥2.10 for Disengagement, which correlates with physical symptoms
to determine the extent of burnout. The OLBI offers advantages over the commonly used Maslach Burnout Inventory, as it uses both positively and negatively framed questions for each domain, which reduces the risk of artefactual relationships
Safety culture perceptions of pharmacists in Malaysian hospitals and health clinics: A multicentre assessment using the Safety Attitudes Questionnaire.
and consists of questions covering 6 patient safety domains of teamwork climate, safety climate, perceptions of management, job satisfaction, working conditions, and stress recognition. These items are scored on a 5-point Likert scale with options including “Strongly Disagree,” “Disagree,” “Neutral,” “Agree,” and “Strongly Agree.” A higher score reflects better safety attitudes. Although the full SAQ comprises 60 questions, each validated version includes the same 30 core questions with additional relevant questions. As this survey was disseminated to HCWs in both clinical and nonclinical roles, respondents had the option to omit domains that were not applicable to them, as some of these domains only applied to clinical situations. A Safety Culture Score was calculated for each domain
(Mean value of item scores within a domain – 1) × 25
Thus, a score of “1” is transformed to “0,” ”2” to “25,” “3” to “50,” “4” to “75,” and “5” to “100.” A score of ≥75 is a “Percentage Agree” for that domain and a “Percentage Agree Rate” is the proportion of respondents with a “Percentage Agree.” Conversely, a score of ≤50 represented “Percentage Disagree.”
is a self-reported questionnaire evaluating Depression and Anxiety with 7 items each (Supplementary Table 4). Each item is rated on a 4-point Likert scale scored as 0, 1, 2, and 3. A score of ≤7 is normal, 8 to 10 is borderline abnormal, and ≥11 is abnormal. We deemed a score of ≥8 or more to signify risk of depression and/or anxiety.
Outcomes
Our primary outcome measure was OLBI mean scores. Secondary outcomes measured included burnout rates based on OLBI-D ≥2.10 and OLBI-E ≥2.25, SAQ Percentage Agree Rates overall and for each domain and proportion of HCWs with a score of ≥8 for HADS-Anxiety and -Depression.
Statistics
Analyses were performed using SPSS 26.0 (IBM Corp, Armonk, NY) with statistical significance set as P < .05.
Cronbach's alpha was presented to show the internal consistency of each questionnaire where an α > 0.7 suggested good reliability. Confirmatory factor analysis was also performed to assess the goodness of fit of the data on the instruments used. Root mean square error of approximation (RMSEA < 0.06), Comparative Fit Indices (CFI ≥0.90) and Standardized Root Mean Square Residual (SRMSR < 0.08) were presented.
OLBI scores in each of the subscales, that is, disengagement and exhaustion, were used as continuous variables and described using the mean and SD. Crude and adjusted predictors (demographic and workplace characteristics as well as HADS and SAQ domain scores) for the OLBI scores were performed using mixed linear models with institution as a random effect.
Ethics
Waiver of consent and ethics approval was obtained from the National Healthcare Group's Domain Specific Review Board (Reference Number 2020/00598). The questionnaire's front page provided participants with information regarding the purpose of the study and assurance of anonymity.
Funding
No funding was received directly for this study. The authors declare no conflicts of interest.
Results
Survey Responses
Among 11,286 invited HCWs, we received 3075 valid responses, which constituted complete demographic and workplace information, OLBI scores, and HADS scores, giving an overall response rate of 27.2%. Although respondents could omit SAQ domain questions that were not appropriate to their work, 94.3% completed at least 1 SAQ domain and 62.7% completed all 6 domains. Table 1 shows the demographic characteristics. Women comprised 71.5% and HCWs of Chinese ethnicity comprised 53.3% with the remainder being of Malay, Indian, and Other ethnicities in roughly equal proportions. Doctors, nurses, AHPs, support staff, and administrative staff comprised 14.9%, 45.3%, 15.7%, 16.0%, and 8.0% of respondents, respectively, with response rates within each HCW role of 38.6%, 31.1%, 23.1%, 23.4%, and 17.5%, respectively.
Table 1Respondent Demographics, Work Environment Characteristics, and Baseline Measures of Emotional Well-Being (n = 3075)
Internal Consistency and Internal Construct Validity
In our study, the Cronbach's alpha for each subscale was good to excellent: OLBI-Exhaustion (α = 0.80) and Disengagement (α = 0.83), HADS Depression (α = 0.80) and Anxiety (α = 0.84), and SAQ Teamwork (α = 0.86), Safety Culture (α = 0.83), Job Satisfaction (α = 0.91), Perceptions of Management (α = 0.83), Stress Recognition (α = 0.83), and Working Conditions (α = 0.83). Goodness of fit indices for most subscales demonstrated a good model fit (Supplementary Table 5).
Scale Scores
In our study population, the scores for each scale are seen in Table 1. The average OLBI scores were 2.38 and 2.50 for Disengagement and Exhaustion, respectively. Burnout thresholds were met by 79.7% and 75.3% of respondents for Disengagement and Exhaustion, respectively, with 86.8% meeting thresholds for either and 68.2% for both. The mean Disengagement scores were highest for administrative staff (2.46) and lowest for support staff (2.32), whereas mean Exhaustion scores were highest for nurses (2.52) and lowest for support staff (2.44), although there was no significant difference in scores among HCW roles (see Table 2 and Figure 1). Mean HADS Depression and Anxiety scores were 5.7 and 6.9, respectively. Average Total SAQ Percentage Agree Rate was 25.9% with the lowest domain being Stress Recognition (8.2%) and highest, Teamwork (55.9%).
Table 2Multivariate Analysis for Total Study Population (n = 3075) Using OLBI-Disengagement and -Exhaustion Scores as Dependent variables
In community refers to Foreign worker dormitories, Community Care Facilities, or Swab Isolation Facilities.
100 (3.26)
2.28 (0.45)
ref
2.40 (0.47)
ref
Work from home
201 (6.54)
2.40 (0.46)
−0.06 (−0.16 to 0.09)
.275
2.48 (0.50)
−0.07 (−0.17 to 0.03)
.194
Tested for COVID-19
Yes
527 (17.1)
2.33 (0.43)
ref
2.47 (0.44)
ref
No
2548 (82.9)
2.39 (0.46)
0.04 (−0.003 to 0.07)
.069
2.50 (0.48)
0.03 (−0.01 to 0.07)
.125
Duration of shift
Overall, h
.169
<.001
< 8
299 (9.72)
2.29 (0.43)
ref
2.33 (0.46)
ref
8 to < 12
2462 (80.1)
2.38 (0.45)
0.002 (−0.09 to 0.04)
.915
2.49 (0.46)
0.06 (0.01 to 0.10)
.015
≥ 12
314 (10.2)
2.49 (0.52)
0.04 (−0.002 to 0.09)
.059
2.69 (0.50)
0.16 (0.10 to 0.23)
<.001
Redeployed
Yes
558 (18.1)
2.45 (0.46)
0.08 (0.04 to 0.11)
<.001
2.55 (0.48)
0.04 (0.01 to 0.08)
.020
No
2517 (81.9)
2.37 (0.45)
ref
2.48 (0.47)
ref
Hospital Anxiety and Depression Scale
Depression
Yes: score ≥ 8
979 (31.8)
2.67 (0.48)
0.19 (0.15 to 0.22)
<.001
2.82 (0.44)
0.23 (0.19 to 0.62)
<.001
No: score <8
2095 (68.2)
2.25 (0.39)
ref
2.34 (0.41)
ref
Anxiety
Yes: score ≥ 8
1253 (40.7)
2.60 (0.45)
0.14 (0.11 to 0.17)
<.001
2.78 (0.42)
0.25 (0.22 to 0.28)
<.001
No: score <8
1822 (59.3)
2.23 (0.40)
ref
2.30 (0.40)
ref
Safety Assessment Questionnaire
Teamwork
Percentage agree
1583 (56.7)
2.22 (0.40)
−0.05 (−0.08 to −0.01)
.013
2.36 (0.43)
−0.02 (−0.06 to 0.01)
.192
Percentage disagree
1208 (43.3)
2.58 (0.46)
ref
2.69 (0.46)
ref
Safety climate
Percentage agree
1512 (53.5)
2.20 (0.38)
−0.04 (−0.08 to −0.01)
.021
2.33 (0.43)
−0.03 (−0.07 to 0.01)
.120
Percentage disagree
1314 (46.5)
2.59 (0.46)
ref
2.70 (0.45)
ref
Job satisfaction
Percentage agree
1715 (53.4)
2.17 (0.35)
−0.28 (−0.31 to −0.24)
<.001
2.32 (0.41)
−0.17 (−0.21 to −0.13)
<.001
Percentage disagree
1173 (40.6)
2.69 (0.43)
ref
2.77 (0.44)
ref
Stress recognition
Percentage agree
228 (7.94)
1.99 (0.39)
−0.20 (−0.25 to −0.15)
<.001
2.05 (0.42)
−0.26 (−0.31 to −0.21)
<.001
Percentage disagree
2642 (92.1)
2.42 (0.45)
ref
2.54 (0.46)
ref
Perception of management
Percentage agree
1059 (36.8)
2.16 (0.39)
−0.07 (−0.11 to −0.04)
<.001
2.27 (0.41)
−0.08 (−0.11 to −0.04)
<.001
Percentage disagree
1819 (63.2)
2.51 (0.45)
ref
2.63 (0.46)
ref
Working conditions
Percentage agree
1232 (44.8)
2.17 (0.37)
−0.07 (−0.10 to −0.03)
<.001
2.29 (0.41)
−0.09 (−0.12 to −0.05)
<.001
Percentage disagree
1517 (55.2)
2.54 (0.46)
ref
2.66 (0.46)
ref
NOTE. Bold values are statistically significant (P < .05).
Mixed Model analysis with Institution as random effect. b = the linear regression estimate which reflects the difference in the OLBI Disengagement or Exhaustion score between groups.
∗ Percentage was calculated as a percentage of valid responses for that covariate and placed in parentheses.
† In community refers to Foreign worker dormitories, Community Care Facilities, or Swab Isolation Facilities.
On mixed model regression analysis with institution as random effects (Table 2), gender, site of work, and being tested for COVID-19 did not have significant differences in OLBI scores, although female gender approached significance for Exhaustion (P = .051). HCWs of Malay and Chinese ethnicities had significantly higher OLBI scores compared with HCWs of Indian ethnicity. Degree holders had significantly higher OLBI scores than HCWs with secondary or lower educational qualifications. Shifts lasting ≥8 hours were associated with significantly higher Exhaustion scores. HADS scores ≥8 for either depression or anxiety were strongly associated (P ≤ .001) with significantly higher OLBI scores. Respondents with a Percentage Agree for any of the SAQ domains had significantly lower Disengagement score. For Exhaustion, this relationship was seen with all SAQ domains except Teamwork and Safety Climate. Redeployment was also associated with significantly higher Exhaustion and Disengagement scores and underwent further subgroup analysis.
Redeployed Subgroup Analysis
A total of 486 (15.8%) HCWs in clinical roles (ie, doctors, nurses, and AHPs) were redeployed. Among redeployed HCWs, the mean Disengagement and Exhaustion scores were 2.45 and 2.55, respectively, both being significantly higher than in non-redeployed HCWs at 2.37 and 2.50 (Table 3), respectively. Among the 3 redeployment groups, that is, (1) Onsite (low risk), (2) Onsite (high risk), and (3) Offsite, a paradoxical but statistically significant relationship was demonstrated in which HCWs redeployed offsite had the lowest Disengagement and Exhaustion scores (2.31 and 2.44) followed by non-redeployed (2.37 and 2.48) and redeployed onsite (low risk 2.49 and 2.62, high risk 2.51 and 2.61). In addition, SAQ Percentage Agree Rates for each domain were highest among HCWs redeployed offsite with significantly higher rates seen for Job Satisfaction and Perceptions of Management. On multivariate analysis, factors that were strongly correlated with high OLBI scores included redeployment onsite (especially high risk), training assessed to be neutral or worse, shift duration ≥8 hours (Table 4).
Table 3Comparison of SAQ, OLBI, and Training Quality in Non-redeployed and Redeployed Clinical HCWs, That is, Doctors, Nurses, Allied Health Professionals
In our sample population, HCWs were redeployed to 1 of 3 areas: (1) within their own work facility with a low risk of COVID-19 contact, that is, Onsite (Low Risk), (2) within their own work facility with a high risk of COVID-19 contact, that is, Onsite (High Risk), or (3) different facility (foreign work dormitory, community care facility, swab isolation facility) with high risk of COVID-19 contact, that is, Offsite.
P Value
Onsite (Low Risk) n = 122
Onsite (High Risk) n = 214
Offsite n = 123
OLBI Score (Mean)
Disengagement
2.37
2.49
2.51
2.31
<.001
Exhaustion
2.50
2.62
2.61
2.44
<.001
SAQ percentage agree rate (%)
Teamwork
57.8
49.6
60.1
65.3
.081
Safety Culture
54.4
44.7
55.6
54.4
.201
Stress Recognition
7.9
3.2
7.6
6.4
.266
Job Satisfaction
58.6
46.4
56.1
72.8
<.001
Perceptions of Management
36.4
29.8
29.1
44.8
.012
Work Culture
42.6
31.5
42.6
45.6
.086
Total
Individual assessment of training quality (%)
Good or better
36.1
47.9
52.8
.002
Neutral or worse
38.5
41.9
32.0
No training received
25.4
10.2
15.2
∗ In our sample population, HCWs were redeployed to 1 of 3 areas: (1) within their own work facility with a low risk of COVID-19 contact, that is, Onsite (Low Risk), (2) within their own work facility with a high risk of COVID-19 contact, that is, Onsite (High Risk), or (3) different facility (foreign work dormitory, community care facility, swab isolation facility) with high risk of COVID-19 contact, that is, Offsite.
Table 4Subgroup Analysis of Redeployed Clinical HCWs (n = 459), That Is, Doctors, Nurses, Allied Health Professionals Using Oldenburg Burnout Inventory (OLBI) Disengagement and Exhaustion Scores as Dependent Variables
In our sample population, HCWs were redeployed to 1 of 3 areas: (1) within their own work facility with a low risk of COVID-19 contact, that is, Onsite (Low Risk), (2) within their own work facility with a high risk of COVID-19 contact, that is, Onsite (High Risk), or (3) different facility (foreign work dormitory, community care facility, swab isolation facility) with high risk of COVID-19 contact, that is, Offsite.
to
.001
.009
Onsite (low risk)
125
2.49 (0.46)
0.18 (0.06 to 0.29)
.004
2.62 (0.47)
0.18 (0.05 to 0.29)
.007
Onsite (high risk)
221
2.51 (0.46)
0.20 (0.09 to 0.30)
<.001
2.61 (0.48)
0.17 (0.05 to 0.28)
.005
Offsite
124
2.31 (0.42)
ref
2.44 (0.48)
ref
Individual assessment of training quality
<.001
<.001
Good or better
210
2.31 (0.41)
ref
2.44 (0.46)
ref
Neutral or worse
177
2.59 (0.46)
0.28 (0.20 to 0.37)
<.001
2.69 (0.49)
0.25 (0.16 to 0.35)
<.001
No training received
72
2.56 (0.47)
0.25 (0.13 to 0.36)
<.001
2.68 (0.48)
0.24 (0.12 to 0.37)
<.001
Tested for COVID-19
Yes
67
2.42 (0.53)
ref
2.57 (0.48)
ref
No
403
2.46 (0.45)
0.04 (−0.10 to 0.16)
.477
2.59 (0.52)
0.02 (−0.16 to 0.10)
.633
Duration of shift, h
.007
< .001
< 8
53
2.27 (0.35)
ref
2.31 (0.43)
ref
8 to < 12
358
2.46 (0.46)
0.19 (0.05 to 0.31)
.007
2.58 (0.47)
0.27 (0.13 to 0.40)
<.001
≥ 12
59
2.55 (0.47)
0.28 (0.09 to 0.43)
.002
2.70 (0.51)
0.39 (0.21 to 0.56)
<.001
NOTE. The bolded values have achieved pre-determined levels statistical significance amongst components of each subcategory.
Mixed model with Institution as random effects performed.
∗ In our sample population, HCWs were redeployed to 1 of 3 areas: (1) within their own work facility with a low risk of COVID-19 contact, that is, Onsite (Low Risk), (2) within their own work facility with a high risk of COVID-19 contact, that is, Onsite (High Risk), or (3) different facility (foreign work dormitory, community care facility, swab isolation facility) with high risk of COVID-19 contact, that is, Offsite.
The few studies conducted on burnout among HCWs in Singapore have mostly used the Maslach Burnout Inventory and note burnout rates ranging from 40% to 60%.
The only pre-pandemic study in Singapore using OLBI involved 37 mental health HCWs and showed mean Exhaustion and Disengagement scores of 2.38 and 2.25,
respectively, which is lower than this study's 2.50 and 2.38. Our study is unique in comparing burnout against SAQ as a surrogate for workplace safety environment during a pandemic among other variables and did so at a timely juncture of 4 months after Singapore's first case and 2 months after instituting major changes to the public health system to combat COVID-19.
Demographic factors that were significantly associated with burnout included ethnicity and educational level. HCWs of Malay and Chinese ethnicities had higher burnout scores than those of Indian or other ethnicities. This has been noted in other studies in Malaysia
where Chinese and Malays constitute the 2 largest ethnic groups. This may be influenced by religio-cultural factors in ways that are not yet fully understood. Higher educational status was associated with higher burnout, as it is likely associated with seniority in health care and thus greater responsibilities. In our study, women had higher Exhaustion scores, which is consistent with other studies
In our study, female-dominated HCW roles included nurses (88.4%), AHPs (73.7%), and administrative (71.1%), whereas doctors and support staff had a roughly equal divide (48.5% and 55.4%, respectively). Nevertheless, after adjusting for factors such as gender, multivariate analysis did not show significant difference in burnout between different HCW roles, although there was a nonsignificant trend toward higher exhaustion scores among nurses that has been observed in other studies.
Surprisingly, HCWs in administrative roles and those who work from home had relatively high disengagement scores, which may be linked to increased operational demands while switching to a different working environment at home. The observation that every strata of the health care workforce can be at risk of increased psychological burden was also noted by Rossi et al.,
who found that nonfrontline HCWs had largely comparable psychological outcomes, such as anxiety, depression, insomnia, and perceived stress levels compared with frontline HCWs.
Various countries have had to redeploy HCWs during this pandemic,
but the effects of redeployment during a pandemic on HCW burnout have not been well studied. The paradoxically higher OLBI scores among HCWs redeployed onsite versus those redeployed offsite and non-redeployed, challenged our hypothesis that unfamiliar work environments under physically demanding conditions
would cause more burnout. Possibly, HCWs redeployed onsite worked with sicker patients compared with those being redeployed offsite within a community setting. Unfortunately, our study did not capture the details and complexities of care within each area of redeployment. A significantly greater proportion of HCWs redeployed offsite rated their training as good or better (52.8%) versus those redeployed onsite (36.1%–47.9%). This was strongly associated with lower OLBI scores and underscores the importance of effective pre-deployment preparation. Many of the HCWs redeployed offsite were volunteers, unlike HCWs who were redeployed onsite, often out of operational necessity or closure of nonessential services. This may contribute to the higher Job Satisfaction and Perceptions of Management Percentage Agree Rates among HCWs redeployed offsite. Altruism
Empathy in clinical practice: How individual dispositions, gender, and experience moderate empathic concern, burnout, and emotional distress in physicians.
noted that the odds ratio of having psychiatric morbidity (defined as “case” under General Health Questionnaire) was highest in unwilling HCWs followed by HCWs without objections and last willing HCWs. Hu et al.,
Frontline nurses’ burnout, anxiety, depression, and fear statuses and their associated factors during the COVID-19 outbreak in Wuhan, China: A large-scale cross-sectional study.
however, noted that although staff dispatched voluntarily from elsewhere to Wuhan had significantly lower emotional exhaustion scores, they had significantly higher depersonalization scores compared with those assigned there. Finally, the strong association between high SAQ Percentage Agree Rates and low OLBI scores highlight the importance of a supportive work safety environment in reducing burnout.
Psychological impact of the 2003 severe acute respiratory syndrome outbreak on health care workers in a medium size regional general hospital in Singapore.
Limitations to this study include the lack of a comparable pre-pandemic health care workforce burnout for direct comparison. Meaningful comparisons between different study populations can also be challenging due to adoption of different rating tools and burnout criteria in different studies and varying cultural acceptance of workplace factors. There also may be a sampling bias, as overworked HCWs may be too busy to respond to this questionnaire. However, we have attempted to mitigate this through 3 rounds of e-mails and managed to obtain a representative cross-section of the health care workforce.
Finally, the citizenship status of HCWs was not captured. In the background of travel bans and quarantine requirements during the pandemic, prolonged time away from family and reduced domestic support may also be an independent risk factor for burnout.
Conclusions and Implications
Our study highlights that every level of the health care workforce is susceptible to burnout. Mitigating strategies should be deployed to both front- and second-line HCWs. Women and HCWs redeployed onsite, especially where involuntary or involving complex medical care, need special attention. This includes female-dominated HCW roles, such as nursing, AHPs, and administrative staff within our health care workforce. Modifiable workplace factors include adequate training, avoiding prolonged shifts ≥8 hours and promoting safe working environments. Future interpandemic strategies include codifying best practices in clinical care and human resource management in preparation for future pandemics as well as continual training and accreditation in infectious disease-relevant skills such as PPE usage. Coping strategies should be taught during and in-between pandemics to reduce the onset and effects of burnout as a continual priority in sustaining patient-care efforts.
Supplementary Data
Supplementary Table 1Health Care Institutions That Were Included in This Study
Health Institution
Address
Bed Capacity
COVID-19 Cases at Any 1 Time During Study Period
Number of Health Care Workers Invited
National University Hospital
5 Lower Kent Ridge Rd, Singapore 119074
1200 beds
50–100
4747
Ng Teng Fong General Hospital
1 Jurong East Street 21, Singapore 609606
700 beds
50–100
2452
Alexandra Hospital
378 Alexandra Rd, Singapore 159964
300 beds
50–100
815
Institute of Mental Health
10 Buangkok View, Buangkok Green Medical Park, Singapore 539747
2000 beds
1–25
2486
National University Polyclinics
Bukit Batok Polyclinic 50 Bukit Batok West Avenue 3 Singapore 659164 Choa Chu Kang Polyclinic 2 Teck Whye Crescent Singapore 688846 Clementi Polyclinic Blk 451 Clementi Avenue 3 #02–307 Singapore 120451 Jurong Polyclinic 190 Jurong East Avenue 1 Singapore 609788 Pioneer Polyclinic 26 Jurong West Street 61 Singapore 648201 Queenstown Polyclinic 580 Stirling Road Singapore 148958
Psychological impact of the 2003 severe acute respiratory syndrome outbreak on health care workers in a medium size regional general hospital in Singapore.
Safety culture perceptions of pharmacists in Malaysian hospitals and health clinics: A multicentre assessment using the Safety Attitudes Questionnaire.
Empathy in clinical practice: How individual dispositions, gender, and experience moderate empathic concern, burnout, and emotional distress in physicians.
Frontline nurses’ burnout, anxiety, depression, and fear statuses and their associated factors during the COVID-19 outbreak in Wuhan, China: A large-scale cross-sectional study.
30835-5&id=gr1.jpg){kind=link}