The Effect of COVID-19 Endemicity on the Mental Health of Health Workers.

OBJECTIVES: A major surge in COVID-19 cases despite Singapore's high vaccination has strained the health care system in October 2021. Our aim was to assess and compare Healthcare Worker (HCW) mental well-being in 2021 against a previously published cohort in 2020.
DESIGN: Cross-sectional survey study. SETTING AND PARTICIPANTS: HCWs from 4 public hospitals and a primary health care system over a 4-week duration in 2021 coinciding with a major surge compared with a similar period in 2020.
METHODS: A survey comprising of the Oldenburg Burnout Inventory (OLBI), Hospital Anxiety and Depression Scale (HADS), and Safety Attitudes Questionnaire (SAQ) was distributed via email. Primary endpoints were the proportion meeting OLBI thresholds for both disengagement and exhaustion and being at risk for both Anxiety and Depression using HADS. Multivariate analysis identified significant predictors among demographic, workplace, and SAQ data. Subgroup analysis of overseas HCWs was performed.
RESULTS: We surveyed 1475 HCWs. Significantly more HCWs met primary outcomes using OLBI and HADS than in 2020 (84.1% and 39.6% vs 68.2% and 23.3%, respectively; P < .001). Burnout levels were uniformly high. A HADS score ≥8 in either subscale was significantly associated with meeting burnout thresholds (P < .001). Overseas HCWs (P = .002), South Asian ethnicity (P = .004), preuniversity educational qualifications (P = .026), and longer shift workhours of 8 to <12 (P = .015) and ≥12 (P = .001) were significantly associated with meeting HADS thresholds. Among overseas HCWs (n=407), seeing family more than a year ago was significantly associated with worse OLBI disengagement scores and a greater proportion meeting HADS thresholds vs seeing them within a year or being local HCWs (47.2% vs 37.2% and 35.6%, respectively; P = .001). CONCLUSIONS AND IMPLICATIONS: HCW mental health has objectively worsened between 2020 and 2021 in the pandemic's second year. Avoiding prolonged shifts, adopting preventive mental health strategies, improving patient safety, and attention to HCWs of minority ethnicity, from overseas, and with preuniversity education may help.

Numerous studies and media have reported high levels of burnout1, 2, 3 and reduced well-being , among health care workers (HCWs) during the COVID-19 pandemic. In 2020, the pandemic was characterized by uncertainty about the disease, high patient loads, , redeployment to unfamiliar and high-risk areas, inadequate personal protective equipment (PPE) , and a lack of robust treatment or preventive options. Reduced physical interaction due to social distancing policies further compounded this by removing the supportive effects of peers and family. This was especially difficult for overseas HCWs, who comprise a large proportion of several countries' health care workforces and are hindered by travel restrictions. , Nevertheless, few studies assessing HCW burnout during this pandemic, as characterized by excessive exhaustion and detachment from one’s job, did not have objective local prepandemic baseline data to definitively conclude causality given that HCWs suffer a high preexisting prevalence of burnout.16, 17, 18 Indeed, mitigating factors, such as altruism, have emerged as themes during this pandemic, although opinions regarding the extent of their protective effects have been conflicting. , ,

Although the second year of the pandemic has been characterized by a better understanding of COVID-19, the production of effective vaccines, a larger armamentarium of treatment options, , and better PPE availability, it has also seen the emergence of more transmissible variants of concern such as the delta , and omicron variants, , which has increased uncertainty. Several countries, such as Singapore, are increasingly transiting from a zero-COVID policy to one that accepts COVID-19 endemicity. Despite our high vaccination rates (>85% of Singapore’s population), several factors such as the emergence of variants of concern, , opening of travel lanes, and outbreaks among vaccinated populations , 32, 33, 34 have contributed to a paradoxical surge in cases since the start of the pandemic from less than 0.7 daily cases per 100,000 population (April to August 2020) to a record high of 58.1 (October 2021) (Figure 1 ). Although about 98% are asymptomatic or mild, serious cases have risen exponentially from less than 0.04 daily deaths per 1,000,000 population to 1.83 and a previous high of 31 critical care cases to a current peak of 67 over the same time frame.

Fig. 1

Multivariate analysis for predictors of meeting thresholds for (1) OLBI disengagement and exhaustion and (2) HADS anxiety and depression.

At present, Singapore’s health care system is strained by a high volume of COVID-19 cases comprising individuals who are unvaccinated or have poor response despite home-based recovery being proposed as the standard of care. This additionally coincides with a return to prepandemic non–COVID-19 caseloads, , which declined in 2020. The chronic, unrelenting volume of cases since the start of the pandemic combined with the continuation of restrictions have likely contributed to an exodus of health care workers from our workforce. ,

As more countries embark on this transition, it will be important to identify contributing factors to HCW mental health, well-being, and burnout in anticipation of possible strains on health care systems. In this follow-up study, we aim to compare the current prevalence of burnout and adverse mental health outcomes in HCWs with that derived from our previous study conducted in 2020 early in Singapore’s COVID-19 experience. Additionally, we aim to understand the effects on our HCWs from overseas.

Methods

We conducted a multicenter, cross-sectional study and disseminated a survey to all HCWs across 4 public hospitals and a network of primary health care centers (polyclinics) between September 19 and October 19, 2021 (Supplementary Table 1). This time frame coincided with the start of a surge in cases despite reaching milestone community vaccination levels. Health care workers surveyed comprised doctors, nurses, allied health care professionals (AHPs), administrative or managerial staff, and support staff. Support staff referred to nonclinical employees involved in hospital operations (eg, porters, housekeeping, and security staff). As cross-sectional studies are unable to assess temporal relationships between exposure and outcome, we sought to mitigate this by comparing results in an earlier study collected over a similar duration from May 29 to June 24, 2020, which coincided with the tail end of the first major surge. The survey was distributed to HCWs in the same health care institutions and via the same methods as the earlier study, that is, through a link to a questionnaire platform (FormSG; GovTech, Singapore) which was disseminated via corporate email and snowball sampling by word of mouth. An email reminder was disseminated twice at the middle and toward the end of each study period. This questionnaire was worded in English, which is Singapore’s lingua franca. The first page declared the questionnaire’s aim, its voluntary and anonymous nature, and that implied consent was associated with proceeding with the survey.

Supplementary Table 1

Health Care Institutions Included in This Study

Health Institution	Address	Bed Capacity	COVID Cases at Time of Study
National University Hospital	5 Lower Kent Ridge Rd, Singapore 1190746779 5555	1239 beds	More than 100
Ng Teng Fong General Hospital	1 Jurong East Street 21, Singapore 6096066716 2000	700 beds	More than 100
Alexandra Hospital	378 Alexandra Rd, Singapore 1599646472 2000	300 beds	50-100
Institute of Mental health	10 Buangkok View,Buangkok Green Medical Park, Singapore 5397476389 2000	2010 beds	24-49
National University Polyclinics	Bukit Batok Polyclinic50 Bukit Batok West Avenue 3Singapore 659164Choa Chu Kang Polyclinic2 Teck Whye CrescentSingapore 688846Clementi PolyclinicBlk 451 Clementi Avenue 3 #02-307Singapore 120451Jurong Polyclinic190 Jurong East Avenue 1Singapore 609788Pioneer Polyclinic26 Jurong West Street 61Singapore 648201Queenstown Polyclinic580 Stirling RoadSingapore 148958		Outpatient services

We collected basic demographic and workplace environment information data and organized respondents by categories such as (1) HCW roles, (2) gender, (3) ethnicity, (4) redeployment outside primary roles, (5) being tested for COVID-19, (6) primary site of work, (7) if work-from-home or working in offsite facilities was part of their job scope, (8) highest educational qualification, (9) number of workdays and average duration of shifts, (10) vaccination status, and (11) if they were HCWs from overseas. For redeployed HCWs, we further categorized them 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).

Responses to three validated questionnaires, that is, Oldenburg Burnout Inventory (OLBI), Safety Attitudes Questionnaire (SAQ), and Hospital Anxiety and Depression Scale (HADS) were collated. These tools have been described in our prior study. The OLBI comprises 16 positively and negatively framed questions rated on a 4-point Likert-type scale that are equally divided into 2 dimensions—exhaustion (which refers to feelings of emptiness, overwork, physical exhaustion, and a strong need for rest) and disengagement (which refers to distancing from the objects and content of one’s work) (Supplementary Table 2). An HCW was deemed to be burnt out if the mean score of both domains exceeded thresholds of ≥2.25 for exhaustion and ≥2.10 for disengagement, which have been correlated with physical symptoms. , The HADS is a widely validated, , self-reported questionnaire evaluating Depression and Anxiety with 7 items for each domain (Supplementary Table 3) rated on a 4-point Likert scale scored as 0, 1, 2, and 3. A score of ≤7 is normal, 8-10 is borderline abnormal, and ≥11 is abnormal. An HCW with a score ≥8 in both domains is deemed to be at risk of both depression and anxiety. The SAQ (Supplementary Table 4) covers 6 patient safety domains (ie, teamwork climate, safety climate, perceptions of management, job satisfaction, working conditions, and stress recognition) rated on a 5-point Likert-type scale comprising 30 core questions common to all versions of the SAQ and some additional relevant questions from the full SAQ bank. Nonclinical HCWs had the option to omit domains that were not relevant. A Safety Culture Score , for each domain was calculated by the formula (Mean value of item scores within a domain – 1) × 25.

Supplementary Table 2

Oldenburg Burnout Inventory (4-Point Likert Scale)

	Question	Domain	Reversed Questions
1.	I always find new and interesting aspects of my work.	D	✓
2.	There are days when I feel tired before I arrive at work.	E
3.	It happens more and more often that I talk about my work in a negative way.	D
4.	After work, I tend to need more time than in the past in order to relax and feel better	E
5.	I can tolerate the pressure of my work very well.	E	✓
6.	Lately, I tend to think less at work and do my job almost mechanically.	D
7.	I find my work to be a positive challenge.	D	✓
8.	During my work, I often feel emotionally drained.	E
9.	Over time, one can become disconnected from this type of work.	D
10.	After working, I have enough energy for my leisure activities.	E	✓
11.	Sometimes I feel sickened by my work tasks.	D
12.	After my work, I usually feel worn out and weary.	E
13.	This is the only type of work that I can imagine myself doing.	D	✓
14.	Usually, I can manage the amount of my work well.	E	✓
15.	I feel more and more engaged in my work.	D	✓
16.	When I work, I usually feel energized.	E	✓

D, Disengagement; E, Exhaustion.

Supplementary Table 3

Hospital Anxiety and Depression Scale

	Question	Domain	Responses & Score
1.	I feel tense or ‘wound up’	A	Most of the time	A lot of the time	From time to time, occasionally	Not at all
			3	2	1	0
2.	I still enjoy the things I used to enjoy	D	Definitely as much	Not quite so much	Only a little	Hardly at all
			0	1	2	3
3.	I get a sort of frightened feeling as if something awful is about to happen	A	Very definitely and quite badly	Yes, but not too badly	A little, but it doesn’t worry me	Not at all
			3	2	1	0
4.	I can laugh and see the funny side of things	D	As much as I always could	Not quite so much now	Definitely not so much now	Not at all
			0	1	2	3
5.	Worrying thoughts go through my mind	A	A great deal of the time	A lot of the time	From time to time, but not too often	Only occasionally
			3	2	1	0
6.	I feel cheerful	D	Not at all	Not often	Sometimes	Most of the time
			0	1	2	3
7.	I can sit at ease & feel relaxed	A	Definitely	Usually	Not often	Not at all
			0	1	2	3
8.	I feel as if I am slowed down	D	Nearly all the time	Very often	Sometimes	Not at all
			3	2	1	0
9.	I get a sort of frightened feeling like ‘butterflies’ in the stomach	A	Not at all	Occasionally	Quite often	Very often
			0	1	2	3
10.	I have lost interest in my appearance	D	Definitely	I don’t take as much care as I should	I may not take quite as much care	I take just as much care ever
			3	2	1	0
11.	I feel restless as I have to be on the move	A	Very much indeed	Quite a lot	Not very much	Not at all
			3	2	1	0
12.	I look forward with enjoyment to things	D	As much as I ever did	Rather less than I used to	Definitely less than I used to	Hardly at all
			0	1	2	3
13.	I get sudden feelings of panic	A	Very often indeed	Quite often	Not very often	Not at all
			3	2	1	0
14.	I can enjoy a good book or radio or TV program	D	Often	Sometimes	Not often	Very seldom
			0	1	2	3

D: Depression, A: Anxiety.

0-7: Normal, 8-10: Borderline abnormal, 11-21: Abnormal.

Supplementary Table 4

Safety Attitudes Questionnaire (5-Point Likert Scale)

	Question	Domain	Reversed Questions
1.	Nurse input is well received in this clinical area.	TW
2.	In this clinical area, it is difficult to speak up if I perceive a problem with patient care.		✓
3.	Disagreements in this clinical area are resolved appropriately (i.e., not who is right, but what is best for the patient).
4.	I have the support I need from other personnel to care for patients.
5.	It is easy for personnel here to ask questions when there is something that they do not understand.
6.	The physicians and nurses here work together as a well-coordinated team.
7.	I would feel safe being treated here as a patient.	SC
8.	Medical errors are handled appropriately in this clinical area.
9.	I know the proper channels to direct questions regarding patient safety in this clinical area.
10.	I receive appropriate feedback about my performance.
11.	In this clinical area, it is difficult to discuss errors.		✓
12.	I am encouraged by my colleagues to report any patient safety concerns I may have.
13.	The culture in this clinical area makes it easy to learn from the errors of others.
14.	I like my job.	JS
15.	Working here is like being part of a large family.
16.	This is a good place to work.
17.	I am proud to work in this clinical area.
18.	Morale in this clinical area is high.
19.	When my workload becomes excessive, my performance is impaired.	SR	✓
20.	I am less effective at work when fatigued.		✓
21.	I am more likely to make errors in tense or hostile situations.		✓
22.	Fatigue impairs my performance during emergency situations (e.g. emergency resuscitation, seizure).		✓
23.	Management supports my daily efforts.	PM
24.	Management doesn’t knowingly compromise patient safety.
25.	Management is doing a good job.
26.	Problem personnel are dealt with constructively by our management.
27.	I get adequate, timely info about events that might affect my work, from management.
28.	The levels of staffing in this clinical area are sufficient to handle the number of patients.	WC
29.	This hospital does a good job of training new personnel.
30.	All the necessary information for diagnostic and therapeutic decisions is routinely available to me.
31.	Trainees in my discipline are adequately supervised.
32.	I experience good collaboration with nurses in this clinical area.	No domain
33.	I experience good collaboration with staff physicians in this clinical area.
34.	I experience good collaboration with pharmacists in this clinical area.
35.	Communication breakdowns that lead to delays in delivery of care are common.		✓
36.	My suggestions about safety would be acted upon if I expressed them to management.

JS, Job satisfaction; PM, Perceptions of management; SC, Safety culture; SR, Stress recognition; TW, Teamwork; WC, Working conditions.

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 (PAR) is the proportion of respondents with a percentage agree.

Subgroup analysis was performed on HCWs from overseas as these comprised about 20% of the health care workforce in Singapore. As overseas travel would mandate prolonged quarantining on return, this was disallowed unless special permission was given to preserve manpower for surge capacity, expansion of COVID treatment facilities, and higher operational tempo. We hypothesized that prolonged separation from family would have adverse mental health effects in this subgroup. We stratified HCWs from overseas based on whether they had seen their families within the last 1 year.

Outcomes

Our primary outcome measures were the proportions of HCWs meeting thresholds for both OLBI exhaustion and disengagement as well as HADS Anxiety and Depression. This recognizes that HCWs meeting thresholds for both domains in each scale likely demonstrated the highest burnout and adverse mental health risks. Secondary outcomes measured included mean HADS and OLBI scores for each domain as well as SAQ PARs overall and for each domain.

Statistics

All analyses were performed using Stata, version 17.0 (StataCorp LLC) with statistical significance set as 2-sided P <.05. Cronbach alphas and confirmatory factor analysis (utilizing root mean square error of approximation, comparative fit indices, and standardized root mean square residual) demonstrated very good reliability and good model fit for most subscales (Supplementary Table 5).

Supplementary Table 5

Cronbach’s Alpha and Goodness-of-Fit Confirmatory Factor Analysis

Subscale	Cronbach Alpha (>0.7)	Confirmatory Factor Analysis Model-Fit Indices
		CFI (>0.9)	RMSEA (<0.06)	SRMSR (<0.08)
HADS
Depression	0.830	0.976	0.061	0.029
Anxiety	0.865	0.982	0.059	0.024
OLBI
Disengagement	0.825	0.906	0.111	0.055
Exhaustion	0.833	0.929	0.097	0.057
SAQ
Teamwork	0.859	0.994	0.038	0.016
Safety climate	0.812	0.988	0.047	0.021
Job satisfaction	0.892	0.981	0.105	0.025
Stress recognition	0.843	0.976	0.139	0.028
Perception of management	0.801	0.995	0.043	0.015
Working condition	0.768	0.990	0.078	0.019

CFI, comparative fit indices; HADS, Hospital Anxiety and Depression Scale; OLBI, Oldenburg Burnout Inventory; RMSEA, root mean square error of approximation; SAQ, Safety Attitudes Questionnaire; SRMSR, standardized root mean square residual.

Multivariate logistic regressions and generalized linear models were used to evaluate for differences in binary and numerical outcomes, respectively, between the 2021 and 2020 cohorts.

Predictors (demographic, workplace characteristics, and SAQ domain scores) for meeting thresholds for both OLBI exhaustion and disengagement domains as well as both HADS anxiety and depression were investigated using generalized linear mixed model with institution as a random effect and presented as adjusted odds ratios (ORs) with 95% CIs. As respondents comprised random samples from each institution, “Institution” was considered as a random factor (clustering effect) to account for the interinstitutional variance contribution and its impact on the P values. Forest plots are shown for the multivariate analyses (Figure 1). A subgroup analysis on HCWs from overseas has been performed (Table 1 ).

Table 1

Comparison of OLBI, HADS, and SAQ Among Overseas HCWs

	Local HCWs (n = 1068)		Overseas HCWs
			Last Seen Family Within Last Year (n = 45; 11.1%)		Last Seen Family Beyond a Year Ago (n = 362; 88.9%)
	Meeting Threshold (%)	Mean (SD)	Meeting Threshold (%)	Mean (SD)	Meeting Threshold (%)	Mean (SD)
OLBI Score
Disengagement	89.9	2.64 (0.51)∗	82.2	2.50 (0.46)∗	92.3	2.70 (0.50)∗
Exhaustion	88.0	2.77 (0.49)	91.1	2.71 (0.43)	91.2	2.83 (0.48)
Disengagement and exhaustion	83.3		80.0		86.7
HADS
Depression	44.3∗	7.06 (4.23)	46.7∗	7.00 (3.58)	52.8∗	7.65 (3.72)
Anxiety	54.9†	8.50 (4.43)	64.4†	8.58 (3.35)	66.3†	9.11 (3.98)
Depression and anxiety	37.2∗		35.6∗		47.2∗

P < .05.

P < .001.

Ethics

Ethics approval with waiver of individual participant consent was obtained from the National Healthcare Group’s Domain Specific Review Board (reference number 2020/00598).

Results

Survey Responses

We had 1475 respondents (response rate 14.5%) (Table 2 ), with an average age of 36.4 years. Among these, 74.6% were female (1.91% were pregnant) and 67.9% were of Chinese ethnicity (the remainder were of Malay, South Asian, and Other ethnicities in roughly equal proportions). Doctors, nurses, AHPs, support staff, and administrative staff comprised 21.6%, 41.6%, 18.2%, 9.4%, and 9.1% of respondents, respectively. Further, 89.4% worked in a hospital, with the remainder working in a primary care setting. Redeployed HCWs accounted for 14.8% of respondents whereas HCWs from overseas comprised 27.6% of respondents. For highest educational qualification, degree holders and above, preuniversity (ie, associate degrees, diploma, A levels, technical college), and high school and below comprised 80.7%, 15.3%, and 4.0%, respectively. We received valid OLBI and HADS scores for all 1475 respondents, whereas 979 (66.3%) HCWs completed all 6 domains of the SAQ.

Table 2

Respondent Demographic, Work Environment Characteristics, and Baseline Measures of Emotional Well-Being

	2021, n (%)	2020, n (%)	P
Total	1475	3075
Age, y, mean (SD)	36.37 (10.04)	36.94 (9.95)	.29
Gender			.004
Male	321 (21.8)	794 (25.8)
Female	1100 (74.6)	2199 (71.5)
Not stated	54 (3.7)	82 (2.7)
Ethnicity			<.001
Chinese	1001 (67.9)	1608 (52.3)
Malay	120 (8.1)	401 (13.0)
South Asian	128 (8.7)	410 (13.3)
Others, eg, Filipino, Burmese or from Myanmar, Eurasian, Caucasian, Vietnamese	158 (10.7)	506 (16.5)
Not stated	68 (4.6)	159 (4.9)
Current primary place of work
Hospital	1318 (89.4)	2841 (92.4)	<.001
Polyclinic or general practice	157 (10.6)	228 (7.41)
Not stated	0 (0)	6 (0.20)
Role			<.001
Doctor	318 (21.6)	458 (14.9)
Nurse	614 (41.6)	1394 (45.3)
Allied health professional	269 (18.2)	483 (15.7)
Support staff, ie, cleaners, porters, technicians, security	138 (9.4)	491 (16.0)
Administrative and managerial	134 (9.1)	247 (8.0)
Not stated	2 (0.1)	2 (0.1)
Highest education qualification			<.001
Degree and above	1190 (80.7)	2132 (69.3)
Preuniversity (ie, associate degrees, diploma, A levels, technical college)	226 (15.3)	699 (22.7)
High school or below	59 (4.0)	244 (8.0)
Redeployment to other areas			<.001
Not redeployed	1257 (85.2)	2534 (82.4)
Redeployed	218 (14.8)	541 (17.6)
Within hospital, low risk	81 (37.2)	140 (25.9)
Within hospital, high risk	78 (35.8)	245 (45.3)
Outside of hospital	43 (19.7)	156 (28.8)
Not specified	2 (0.92)	0 (0)
Have you ever been infected with COVID-19?			N/A
Yes	17 (1.2)	Not Surveyed
No	1458 (98.8)
COVID-19 status in the last 1 wk			<.001
Negative	1400 (94.9)	506 (16.5)
Positive	9 (0.6)	21 (0.7)
Do not know or not tested	66 (4.5)	254 (82.9)
Vaccination status		Not Applicable	N/A
Completed 2 doses at least	1456 (98.7)
Completed only 1 dose	9 (0.61)
Not vaccinated	10 (0.68)
Have you treated a COVID-19–positive patient in the past week?			<.001
Yes	332 (22.5)	462 (15.0)
No	1059 (71.8)	2449 (79.6)
Don’t know	84 (5.7)	164 (5.3)
Average duration of your shift or daily work hours?			<.001
<8 h	69 (4.7)	299 (9.7)
8-12 h	1248 (84.6)	2462 (80.1)
≥12 h	158 (10.7)	314 (10.2)

Thresholds for deeming burnout for (1) OLBI disengagement ≥2.10 and (2) OLBI exhaustion ≥2.25.

Thresholds for deeming risk of anxiety or depression for HADS was ≥8 in either subscale.

PAR refers to the proportion of respondents who scored 75% or above for the Safety Culture Score in each domain.

Comparison Between 2021 and 2020 Cohorts

Compared to the previous study in 2020 (3075 respondents, response rate 27.2%) (Table 2), there was no significant difference in age (mean age 36.37 vs 36.94, P = .29). However, the 2021 cohort had a slightly higher proportion of females (74.6% vs 71.5%, P = .004). Additionally, there were significant differences observed in ethnic composition, primary place of work, HCW role, highest education qualification, redeployment status, latest COVID-19 testing status, recently treating a COVID-19 patient and shift duration (all P <.001).

In the current 2021 study, the average OLBI disengagement and exhaustion scores were 2.65 and 2.78, respectively, with 90.2% and 88.9% reaching threshold in each domain and 84.1% reaching thresholds in both domains. This was significantly higher than in the 2020 cohort, where 68.2% met burnout threshold in both domains (P <.001). The HADS scores also showed a significantly higher proportion of respondents at risk of mental health issues in 2021, with 39.6% meeting thresholds for risks of both anxiety and depression vs 23.3% in 2020 (P <.001). Finally, SAQ scores also showed a significantly poorer PAR in each domain in 2021, with only 16.9% reaching PAR for the total score compared with 25.9% in 2020 (P <.001).

Multivariate Analysis

In the multivariate analysis (Figure 1), we identified having a HADS score of 8 or more for either the anxiety (OR 4.85, 95% CI 2.72-8.65; P = .001) or depression domains (OR 4.85, 95% CI 2.35-10.0; P <.001) to be independently associated with meeting the thresholds for both OLBI disengagement and exhaustion. Furthermore, reaching a percentage agree in the SAQ domains for job satisfaction (OR 0.13, 95% CI 0.07-0.25; P < .001) and stress recognition (OR 0.25, 95% CI 0.09-0.75; P = .014) was negatively associated with meeting burnout thresholds.

For being at risk of both anxiety and depression, independent predictors included being HCWs of South Asian ethnicity (OR 2.13, 95% CI 1.28-3.55; P = .004), having a preuniversity educational certificate as the highest qualification (eg, diploma, A levels, technical college) (OR 3.53, 95% CI 1.16-10.7; P = .026), being an HCW from overseas (OR 1.74, 95% CI 1.23-2.45; P = .002), and working longer workhours per shift (8 to <12 hours: OR 3.21, 95% CI 1.26-8.18; P = .015; and ≥12 hours: OR 7.40, 95% CI 2.65-20.63; P = .001). In the SAQ, having a percentage agree for safety climate (OR 0.68, 95% CI 0.47-0.98; P = .041), job satisfaction (OR 0.41, 95% CI 0.28-0.59; P <.001), stress recognition (OR 0.17, 95% CI 0.04-0.79; P = .024), and working conditions (OR 0.55, 95% CI 0.36-0.84; P = .006) were negatively associated.

Health Care Workers From Overseas

Among 407 HCWs from overseas (Table 1), 362 (88.9%) had last seen their families more than a year ago. OLBI disengagement scores for this group were significantly higher (2.70) compared with local HCWs and HCWs from overseas who had seen their families within a year ago (2.64 and 2.50, respectively) (P = .023), whereas the percentage meeting thresholds for being at risk of depression and anxiety was higher at 47.2% vs 37.2% and 35.6%, respectively (P = .0013). There was no significant difference, however, in SAQ PARs, mean HADS scores, or the proportion meeting OLBI thresholds for both domains.

Discussion

Our study uniquely addresses the longitudinal impacts of a prolonged pandemic posture on HCW mental well-being by comparing outcomes in 2020 and 2021, in the face of a spike of new cases despite high population vaccination rates. We demonstrated that OLBI and HADS mean scores and proportion of HCWs meeting thresholds for burnout and being at risk of both anxiety and depression has increased significantly between 2020 and 2021. The absolute percentage of HCWs meeting these thresholds in 2021 is at a worrying level of 84.1% and 39.6% respectively as opposed to 68.2% and 23.3% in 2020. Additionally, attitudes toward patient safety have also deteriorated, with a significantly lower PAR for the total score (16.9% vs 25.9% in 2020). This is despite increased awareness and the institution of several protective policies , 49, 50, 51, 52, 53 over the last 2 years.

When comparing independent predictors for burnout between our 2020 and 2021 cohort, only reaching threshold HADS scores for being at risk of depression and/or anxiety was common. Other predictors identified in 2020, such as being redeployed, and longer working hours per shift were no longer seen as independent predictors in our present cohort although this is likely due to uniformly high levels of burnout among different subgroups of our study population, limiting the power to detect significant differences.

Regarding adverse mental health outcomes, South Asian ethnicity (a minority ethnic group in Singapore), longer workhours per shift, holding a preuniversity certificate as the highest education qualification, and being an HCW from overseas were independently associated with risk of anxiety and depression. Although some studies have identified people of minority ethnicity as being at risk for mental health issues, , there is conflicting evidence owing to poorly understood religio-cultural aspects, the role of community support, varied minority experiences in different countries, and significant intraethnic heterogeneity. Indeed, in Singapore’s multiethnic context, minorities in non-HCW populations have frequently demonstrated better self-reported mental health. , This has been postulated to be due to stronger family, religious, and community networks and contrasts against collectivist cultures with Confucian roots among those of Chinese ethnicity that prioritize social and familial harmony over open displays of emotions. This reversal of trends may be due to negative impacts of the pandemic on important supportive community, religious, familial, and social networks. This is an area that warrants further study.

Achieving a percentage agree in several SAQ domains had a statistically significant negative association with the primary outcomes. Although the causal relationship between perceptions of patient safety and burnout, depression, or anxiety is unclear, institutional focus on improving patient safety in deficient domains of the SAQ may be protective. The relationship between long working hours and adverse mental health outcomes is well understood.60, 61, 62 Like our previous study, shifts lasting less than 8 hours may be protective against burnout, highlighting the importance of proper work-rest cycles and prevention of prolonged shift lengths as a potential way to prevent burnout.

Subgroup analysis of HCWs from overseas demonstrated a significant association between prolonged separation from families for longer than a year and higher HADS anxiety and depression mean scores. This supplements the scarce data pertaining to this important and vulnerable group that comprises a significant component of several countries’ HCW workforces. , , Finding ways to allow HCWs from overseas to travel by facilitating and prioritizing leave to return home, enabling more efficient quarantining processes (if required), ensuring completion of vaccination, and expanding psychological support may be beneficial.

A very large proportion of our sample was fully vaccinated (98.7%) or undergoing regular SARS-CoV-2 testing (95.5%). This may be attributed to both of these being mandatory workplace requirements in the absence of valid contraindications, high levels of vaccine acceptance in Singapore, as well as a precedent for mandatory hepatitis B vaccinations and opt-out semiannual influenza vaccination exercises among HCWs. Nevertheless, this reduced the power of our study to determine the effects of these variables on our primary outcomes. The effects of these potentially protective factors should be further evaluated in settings with less HCW vaccination coverage or testing uptake either due to a lack of availability or personal preferences.

Despite an abundance of recommendations on how to combat burnout and adverse mental health outcomes among HCWs, there is limited evidence on their efficacy in this pandemic. , The objective worsening of these outcomes highlights the need for renewed efforts to combat this. This may be through the formal creation of well-resourced internal mental well-being departments dedicated to HCWs at an institutional level with appointed welfare champions spread across departments and health care roles. Our study identified HCWs from overseas and of minority ethnicity as groups vulnerable to adverse mental health outcomes, which we postulate may be due to effects of the pandemic on community support structures. As poorer HADS scores was significantly associated with higher risks of burnout, ensuring good HCW mental health may be protective and this can be done by increasing access to trained counseling services, encouraging regular mental well-being breaks, disseminating reminders about self-care with emphasis on adequate rest, sleep, recreation, and maintaining connections with loved ones amid hectic schedules. , Policies should be enacted or reinforced at the institutional, state, and national level to guarantee the protection of rest cycles, leave taking, child and family care aspects, and manpower allocations. Factors contributory or indicative of work stress such as shift durations, patient numbers, leave taken, days off, and sick leaves should be tabulated with regular mental health surveys. Redline thresholds for escalation to senior management should be assigned. Finally, targeting deficient patient safety domains as guided by SAQ may complement the above recommendations.

Limitations and Strengths

Although the 2021 study adopted the same methodology and recruitment strategy as the 2020 study over a similar study duration, we unexpectedly had fewer respondents (1475 vs 3075). This may have led to a sampling bias with some differences in baseline characteristics of the respondents. Nonetheless, the distribution of respondents’ age was similar, and these differences were adjusted for when performing multivariate analysis to study the primary outcomes. Furthermore, we sampled the same population that would have faced similar changes in workplace conditions and policies between the 2 study periods. Thus, although a similar response rate with equivalent demographic factors would have been ideal, we feel that our findings remain valid. Various factors may have contributed to our reduced response rate. We postulate that HCWs had reduced ability to respond to survey invitation as workloads in 2021 were significantly higher because of a large surge in COVID-19 cases coupled with a return to prepandemic levels of non-COVID cases. , Significantly more HCWs treated COVID-19–positive patients in our present study (22.5% vs 15.0%). Additionally, because of difficulties with predicting case trajectories, the 2020 study coincided with the tail end of the initial surge whereas the 2021 one coincided with the peak of a much larger one (Figure 2 ).

Fig. 2

Daily new cases and percentage of population fully vaccinated in Singapore.

The majority of studies in the current literature that evaluate mental health outcomes in HCWs during the COVID-19 pandemic are cross-sectional in nature, hence limiting their interpretation of causality. We believe a significant strength in our study is that we attempted to mitigate this by comparing the outcomes with our 2020 cohort. Nevertheless, longitudinal follow-up with individual subjects as part of a prospective cohort study design would have improved our understanding of how this relationship has changed with time. Additionally, mixed methods analysis can be employed to qualitatively understand other relevant factors contributing to these worrying trends.

Conclusions and Implications

Singapore’s paradoxical peak in COVID-19 cases despite high population vaccination rates is a cautionary message to all countries aiming to open their economies and accept COVID-19 endemicity. Furthermore, until global vaccine equality is achieved, the risk of future SARS-CoV-2 variants, such as omicron, exists and guarantees that COVID-19 will remain a long-term public health concern. , Although HCW burnout and mental health issues have been a perennial problem, it has worsened over this pandemic despite increased awareness and support. Thus, as we transit to a state of COVID endemicity, we may face a second mental health epidemic among HCWs. We urge health care systems to comprehensively audit the state of mental health among their HCWs and to continually evaluate for contributory factors to guide formal steps toward protecting them and keeping health care delivery sustainable.