Keywords
healthcare workers - sleep quality - insomnia - stress - COVID-19 - Latin America
Introduction
The COVID-19 pandemic has had an immense effect on the mental health and sleep patterns
of healthcare workers (HCWs) worldwide. Many studies have demonstrated the effect
that a pandemic can have on the anxiety, depression, and stress levels of HCWs.[1]
[2] A study performed after the 2003 SARS outbreak in China found that 10% of the HCWs
surveyed had a high intensity of post-traumatic stress symptoms related to their work
in patient care and that the presence of those symptoms was directly proportional
to their exposure to the pandemic.[1]
[2]
It has been recognized that HCWs tend to have poor sleep quantity and quality related
to long work schedules and shiftwork.[3]
[4] This conveys a higher risk of work-related accidents and compromises their general
and mental health.[5] From the Hubei province in China to Europe, studies have been performed to assess
the mental and sleep-related effects the burden of the pandemic has imposed on HCWs
and the function that social support and self-efficacy might have in sleep quality,
depression, and anxiety symptoms during these demanding times worldwide.[6]
[7]
[8]
[9]
[10]
One of the studies was performed in the Hubei province, China during the first wave
of the pandemic; in this survey, 34% of the participants (702 doctors and 1,128 nurses)
had high rates of insomnia measured with the Insomnia Severity Index (ISI). Half of
the participants had depressive symptoms (PHQ9) and 70% reported high-stress levels.
The severity related factors were being female, nurses, and frontline COVID-19 HCWs.[11]
It is important to study the impact the COVID-19 pandemic has on healthcare providers.
Acknowledging these phenomena could identify intervention scenarios that reduce the
emotional burden and promote better patient care. To date, studies in Latin America
have focused mainly on the impact on clinical practice by physicians, illustrating
issues such as the reduction in the number of procedures, face-to-face visits, and
salaries; the implementation of telehealth; or the impact on clinical trials.[12]
[13]
[14]
[15] Nonetheless, the effect of the pandemic on the sleep or mental health of healthcare
workers in this region remains to be elucidated.
We aimed to describe the sleep quality, the frequency of insomnia symptoms, and the
perceived stress during the first wave of the COVID-19 pandemic in healthcare workers
in a high complexity hospital located in Bogotá, Colombia.
Material and Methods
Design and Participants
We conducted a descriptive, observational, cross-sectional study between September
4th and October 20th, 2020, during the first wave of the COVID-19 pandemic in healthcare workers, including
medical staff and residents, nurses, radiology technicians, professionals in respiratory,
physical, and speech therapy, bacteriologists, and patient transport personnel within
a high complexity hospital located in Bogotá, Colombia. No sample size was calculated,
as the study was developed as a census.
Applied Instruments
Initially, we asked participants their age, sex, marital status, school level, profession,
medical specialty (for residents and specialists), work area, years of professional
experience, and whether they participated in clinical activities regarding the diagnosis,
treatment, or care of patients with suspicion or diagnosis of COVID-19. This last
question was made to determine whether the HCWs were at the front line of the pandemic.
This approach is like the one used by Zhang et al.[11] In addition, based on previous reports,[15]
[16] we included a Likert-scale question on the perceived risk of the respondents and
their family members of becoming infected with SARS-CoV-2; we stratified the answers
into low, moderate, and high.
Sleep quality was determined using the Colombian-validated version of the Pittsburgh
Sleep Quality Index (PSQI)[17] which indicated the participants with good or poor sleep quality. Participants with
a PSQI score > 5 points were included in the poor sleep quality group. Participants
with a PSQI score < 5 points were included in the good sleep quality group.
Insomnia was assessed using the Spanish-validated version of the Insomnia Severity
Index (ISI).[18] The ISI is composed of seven items that evaluate the difficulties for falling or
maintaining sleep, early awakenings, degree of satisfaction with current sleep, and
the interferences of poor sleep with normal daytime functioning. For the interpretation
of the ISI, we used a score > 14 to determine the presence of insomnia, which is consistent
with other studies developed during the COVID-19 pandemic.[11]
[19] Additionally, the ISI has been deemed as sensitive to change after cognitive-behavioral
or pharmacological therapies have been instated.[18]
Regarding perceived stress, the 10-item Perceived Stress Scale (PSS-10) Colombian
validated version[20] was used to evaluate the general psychological response to stressors.[21] It evaluates two dimensions: general stress and coping capacities. We considered > 24
points as high perceived stress. This cutoff point has been used previously in studies
performed in Colombia, and it has demonstrated an adequate psychometric performance
to evaluate stress during the COVID-19 pandemic.[22]
The described tools were condensed in a questionnaire using RedCap version 10.7.1,
Vanderbilt University, Nashville, United States of America, and a survey link was
obtained and disseminated with the support of the Hospitaĺs Human Resources office.
Additionally, the link was sent via WhatsApp groups of which potential participants
could be part. The survey's answers were anonymous.
Data Analysis
The software yielded spreadsheets, which were analyzed using R studio, Posit, Massachusetts,
United States of America. Absolute and relative frequencies were calculated for qualitative
variables and central tendency and dispersion measures were determined for quantitative
variables.
Ethical Considerations
The study protocol was reviewed and approved by the Research and Ethics Committee
of the Hospital and University (FM-CIE-0663–20). The study was classified as minimal
risk research and was conducted in agreement with the Helsinki Declaration and Resolution
008430 of 1993 issued by the Colombian Ministry of Health, thus a waiver for informed
consent was obtained.
Results
From the 3,283 workers affiliated with the Hospital, we obtained data from 1,155 healthcare
personnel (35.18% participation) between September 4th through October 20th, 2020. [Table 1] shows the demographic characteristics of the surveyed population and [Table 2] the participation rates by profession. Remarkably, around half of the medical and
nursing staff participated in the study. Fifty percent of the respondents were between
the 31 and 45 years old, and 76 percent were women. Of the HCWs surveyed, 44% were
nursing staff and 30% were medical personnel. Also, 66.4% of the participants were
directly committed to COVID-19 patients, hence considered frontline COVID-19 HCWs.
Table 1
Demographic characteristics. HCW: Healthcare workers.
|
Demographics
|
n (%)
|
|
Gender
|
|
Male
|
269 (23.3)
|
|
Female
|
883 (76.5)
|
|
Prefer not to answer
|
3 (0.3)
|
|
Age group (years old)
|
|
< 30
|
379 (32.8)
|
|
31–45
|
585 (50.6)
|
|
> 45
|
191 (16.5)
|
|
Marital status
|
|
Single
|
512 (44.3)
|
|
Married/Consensual union
|
577 (50.0)
|
|
Widowed/Divorced
|
66 (5.7)
|
|
Education
|
|
Bachelor / Technician / Technologist
|
426 (36.9)
|
|
Undergraduate
|
255 (22.1)
|
|
Postgraduate
|
345 (29.9)
|
|
Master's degree or PhD
|
129 (11.2)
|
|
Profession
|
|
Medical specialist
|
217 (18.8)
|
|
Resident
|
102 (8.8)
|
|
General practitioner
|
28 (2.4)
|
|
Nurse
|
211 (18.3)
|
|
Nursing assistant
|
308 (26.7)
|
|
Radiology technician
|
10 (0.9)
|
|
Physical therapist
|
16 (1.4)
|
|
Respiratory therapist
|
22 (1.9)
|
|
Speech therapist
|
4 (0.3)
|
|
Bacteriologist
|
42 (3.6)
|
|
Stretcher-bearer
|
5 (0.4)
|
|
Laboratory assistant
|
23 (2.0)
|
|
Others
|
167 (14.5)
|
|
Predominant work area
|
|
Intensive care unit
|
106 (9.2)
|
|
COVID emergency room
|
138 (11.9)
|
|
Non-COVID emergency room
|
53 (4.6)
|
|
COVID inpatient
|
105 (9.1)
|
|
Non-COVID inpatient
|
120 (10.4)
|
|
Telehealth
|
40 (3.5)
|
|
Radiology
|
32 (2.8)
|
|
Surgical theater
|
131 (11.3)
|
|
Others
|
430 (37.2)
|
|
Years in current profession
|
|
Mean (SD)
|
11.2 (10.6)
|
|
Care of COVID-19 patients
|
|
Yes
|
769 (66.6)
|
|
No
|
386 (33.4)
|
Abbreviation: SD, standard deviation.
Table 2
Participation rates by profession.
|
Profession
|
Total HCWs (n)
|
HCWs surveyed (n)
|
Participation rate (%)
|
|
Medical specialist
|
426
|
217
|
50.9
|
|
Resident
|
430
|
102
|
23.7
|
|
General practitioner
|
54
|
28
|
51.9
|
|
Nurse
|
385
|
211
|
54.8
|
|
Nursing assistant
|
778
|
308
|
39.6
|
|
Radiology technician
|
33
|
10
|
30.3
|
|
Physical therapist
|
58
|
16
|
27.6
|
|
Respiratory therapist
|
22
|
22
|
100
|
|
Speech therapist
|
5
|
4
|
80.0
|
|
Bacteriologist
|
46
|
42
|
91.3
|
|
Stretcher-bearer
|
36
|
5
|
13.88
|
|
Laboratory assistant
|
43
|
23
|
53.5
|
|
Others
|
973
|
167
|
17.3
|
Abbreviation: HCW, healthcare worker.
[Table 3] shows the descriptive analysis performed for each of the described conditions, only
full responses of each questionnaire were described. Poor sleep quality, insomnia,
and high perceived stress were found in 74.9, 12.4, and 13.2% of the HCWs surveyed,
respectively. Categories were grouped to facilitate analysis and visualization. Additional
information regarding the sociodemographic characteristics and descriptive analysis
of the 319 (27.9%) medical specialists and residents can be found in the [Supplementary Tables 1] and [2].
Table 3
Descriptive analysis.
|
Sleep quality
|
Insomnia
|
Perceived stress
|
|
Good [0–4] (n = 289)
|
Poor[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
[18]
[19]
[20] (n = 863)
|
No [≤ 15] (n = 981)
|
Yes [> 15] (n = 143)
|
No [≤ 25] (n = 913)
|
Yes [> 25] (n = 152)
|
|
Gender
|
|
Male
|
91 (33.8%)
|
178 (66.2%)
|
235 (88.0%)
|
32 (12.0%)
|
235 (90.4%)
|
25 (9.6%)
|
|
Female
|
198 (22.4%)
|
685 (77.6%)
|
746 (87.0%)
|
111 (13.0%)
|
678 (84.2%)
|
127 (15.8%)
|
|
Age (years old)
|
|
< 30
|
73 (19.4%)
|
304 (80.6%)
|
311 (85.2%)
|
54 (14.8%)
|
303 (89.1%)
|
37 (10.9%)
|
|
31–45
|
157 (26.9%)
|
427 (73.1%)
|
503 (88.1%)
|
68 (11.9%)
|
462 (84.2%)
|
87 (15.8%)
|
|
> 45
|
59 (30.9%)
|
132 (69.1%)
|
167 (88.8%)
|
21 (11.2%)
|
148 (84.1%)
|
28 (15.9%)
|
|
Marital status
|
|
Single
|
106 (20.8%)
|
404 (79.2%)
|
425 (85.9%)
|
70 (14.1%)
|
408 (88.1%)
|
55 (11.9%)
|
|
Married/consensual union
|
170 (29.5%)
|
406 (70.5%)
|
497 (88.1%)
|
67 (11.9%)
|
451 (83.4%)
|
90 (16.6%)
|
|
Widowed/divorced
|
13 (19.7%)
|
53 (80.3%)
|
59 (90.8%)
|
6 (9.2%)
|
54 (88.5%)
|
7 (11.5%)
|
|
Care of COVID-19 patients
|
|
Yes
|
192 (25.1%)
|
574 (74.9%)
|
652 (87.0%)
|
97 (13.0%)
|
610 (85.7%)
|
102 (14.3%)
|
|
No
|
97 (25.1%)
|
289 (74.9%)
|
329 (87.7%)
|
46 (12.3%)
|
303 (85.8%)
|
50 (14.2%)
|
|
Education
|
|
Bachelor / Technician / Technologist
|
99 (23.3%)
|
326 (76.7%)
|
358 (86.7%)
|
55 (13.3%)
|
334 (86.8%)
|
51 (13.2%)
|
|
Undergraduate
|
50 (19.7%)
|
204 (80.3%)
|
207 (83.1%)
|
42 (16.9%)
|
205 (87.2%)
|
30 (12.8%)
|
|
Postgraduate
|
92 (26.7%)
|
252 (73.3%)
|
296 (88.6%)
|
38 (11.4%)
|
270 (84.4%)
|
50 (15.6%)
|
|
Master's degree or PhD
|
48 (37.2%)
|
81 (62.8%)
|
120 (93.8%)
|
8 (6.2%)
|
104 (83.2%)
|
21 (16.8%)
|
|
Profession
|
|
Medical staff
|
115 (33.3%)
|
230 (66.7%)
|
302 (88.6%)
|
39 (11.4%)
|
280 (85.4%)
|
48 (14.6%)
|
|
Nursing staff
|
117 (22.6%)
|
401 (77.4%)
|
437 (87.1%)
|
65 (12.9%)
|
417 (87.8%)
|
58 (12.2%)
|
|
Others
|
36 (19.8%)
|
146 (80.2%)
|
154 (86.5%)
|
24 (13.5%)
|
135 (81.8%)
|
30 (18.2%)
|
|
Therapists
|
9 (21.4%)
|
33 (78.6%)
|
35 (87.5%)
|
5 (12.5%)
|
31 (83.8%)
|
6 (16.2%)
|
|
Laboratory
|
12 (18.5%)
|
53 (81.5%)
|
53 (84.1%)
|
10 (15.9%)
|
50 (83.3%)
|
10 (16.7%)
|
|
Predominant work area
|
|
COVID-19 area
|
77 (22.2%)
|
270 (77.8%)
|
287 (84.9%)
|
51 (15.1%)
|
273 (85.6%)
|
46 (14.4%)
|
|
Non-COVID-19 area
|
106 (31.6%)
|
229 (68.4%)
|
295 (89.4%)
|
35 (10.6%)
|
275 (87.0%)
|
41 (13.0%)
|
|
Other
|
106 (22.6%)
|
364 (77.4%)
|
399 (87.5%)
|
57 (12.5%)
|
365 (84.9%)
|
65 (15.1%)
|
|
Years in current profession
|
|
Mean (SD)
|
12.2 (12.5)
|
10.9 (9.9)
|
11.3 (10.3)
|
10.9 (13.3)
|
11.1 (11.0)
|
11.7 (9.2)
|
|
Perceived personal risk of COVID-19 contagion
|
|
Low
|
60 (33.7%)
|
118 (66.3%)
|
158 (91.3%)
|
15 (8.7%)
|
136 (82.4%)
|
29 (17.6%)
|
|
Moderate
|
129 (26.8%)
|
353 (73.2%)
|
426 (91.0%)
|
42 (9.0%)
|
395 (87.8%)
|
55 (12.2%)
|
|
High
|
100 (20.3%)
|
392 (79.7%)
|
397 (82.2%)
|
86 (17.8%)
|
382 (84.9%)
|
68 (15.1%)
|
|
Perceived family risk of COVID-19 contagion
|
|
Low
|
102 (33.4%)
|
203 (66.6%)
|
278 (92.7%)
|
22 (7.3%)
|
253 (88.2%)
|
34 (11.8%)
|
|
Moderate
|
124 (25.2%)
|
369 (74.8%)
|
431 (90.2%)
|
47 (9.8%)
|
402 (87.8%)
|
56 (12.2%)
|
|
High
|
63 (17.8%)
|
291 (82.2%)
|
272 (78.6%)
|
74 (21.4%)
|
258 (80.6%)
|
62 (19.4%)
|
Abbreviation: SD, standard deviation.
Regarding sleep quality, a higher frequency of poor sleep was observed in females
and young (< 30 years old) workers; also, the HCWs who were single, widowed, or divorced
at the time of the study had poorer sleep quality in comparison with married HCWs.
The prevalence of insomnia evaluated using the ISI with a 14-point cutoff was 12.41%,
the frequency of insomnia was relatively similar in all participant groups except
for the workers who referred a personal or family high-risk perception of COVID-19
contagion.
Concerning perceived stress, the cumulative prevalence was 13.2%. Females were more
stressed than men, and younger workers (< 30 years old) were less stressed than their
elder colleagues. No major differences were found in perceived stress levels between
those who cared for SARS-CoV-2 infected patients and those who did not, as found with
insomnia, those who referred a personal or family high-risk perception of COVID-19
contagion had high perceived stress.
Discussion
Elevated levels of stress, depression, and anxiety have been described in healthcare
workers in association with the COVID-19 pandemic.[11]
[23]
[24]
[25]
[26] The experience with previous pandemics has shown high rates of posttraumatic stress
in relation to patient care. We performed a descriptive, cross-sectional study in
1,155 workers in a high complexity hospital in Latin America during the first wave
of SARS-CoV-2 contagion when the Colombian healthcare system was on the brink of collapse.
To the best of our knowledge, this is the first study published evaluating these three
outcomes in a Latin American HCWs population during the current pandemic. It is also
one of the few studies that include health professionals other than medical and nursing
staff by including therapists and technicians.
The finding that 74.9% of the surveyed healthcare workers had poor sleep quality is
alarming. This is like the findings during SARS in 2003.[2]
[27] As found by other studies, women had worse sleep quality than men.[11]
[19] No statistically significant differences were found between the HCW that were directly
involved in the care of COVID-19 patients (frontline healthcare workers) and those
who did not, which is consistent with studies such as the one performed by Jahrami
in Bahrain.[23]
Our study found a 12.4% frequency of insomnia. This percentage is lower than the one
found by Zhang et al[11] in Wuhan, China (36.1%) and is also lower than the ones found during the SARS outbreak
in 2003 in Hong Kong and Taiwan.[27]
[28] These variations in frequency could be due to the variability of the definition
of insomnia between studies. While others have contemplated a score > 8 in the ISI
as positive for insomnia, our study considered a score > 14; this cutoff point has
been found to have greater clinical significance.[18]
[19] The frequency of insomnia was found to be similar among the sociodemographic variables
analyzed, suggesting that, regardless of the characteristics or profile of the HCW,
insomnia was present in a similar degree. Interestingly, based on the reported self-perceived
risk of personal or the respondent's family members becoming infected with SARS-CoV-2,
those with a low self-perceived risk had a lower prevalence of insomnia.
Regarding perceived stress, resembling what occurred with insomnia, there was a similar
frequency regardless of the contrasted sociodemographic characteristics. There were
only differences between males and females, the latter being the ones with the highest
frequency of perceived stress.
As in other mental health and sleep quality-related studies,[29]
[30]
[31]
[32]
[33] females were at a higher risk of being poor sleepers. Also, family ties (that is,
being married or in a consensual union) implied an association with poor sleep quality.
This concurs with the finding that poor sleep quality was highly present in the moderate
to high perceived risk of COVID-19 contagion by the family members of the workers
surveyed.
Despite the alarming results, our study has some limitations. First, being a cross-sectional
study, the baseline status of the main three phenomena analyzed (sleep quality, insomnia,
and perceived stress) is unknown, as well as the progression with the epidemiological
evolution of infections in the country. Studies performed after the 2003 SARS outbreak
showed a progressive clearing of insomnia symptoms in the two weeks after the end
of the crisis. Nevertheless, the situation with COVID-19 has been different, as subsequent
contagion peaks have appeared over a prolonged period. Two brief follow-up longitudinal
studies have been performed evaluating the evolution of these symptoms over time in
relation to the COVID-19 pandemic, suggesting the need for long-term psychological
and sleep-related support for HCWs.[34]
[35] Second, we did not inquire about shift work nor quantified changes in the workload
of the workers, which could be related to higher frequencies of poor sleep quality,
insomnia, and perceived stress.[36]
[37]
[38]
Conclusion
Our findings call for the development of precise strategies prioritizing those at
greater risk of having poor sleep quality, such as females and married workers. Additionally,
they convey the importance of generating cross-sectional strategies for the management
of insomnia and stress in healthcare workers, regardless of their sociodemographic
profile. Longitudinal studies should be performed in the future to assess the evolution
of these three conditions over time.
