Anxiety and Depression in Psychiatry Residents of Universitas Gadjah Mada During Early COVID-19 Pandemic

Abstract

Health workers at the forefront in handling COVID-19 cases are particularly vulnerable to contracting COVID-19 infections and mental health problems. Psychiatry residents who studied and worked in hospitals were expected to provide mental health and psychosocial treatments to health workers and COVID-19 patients. However, it was necessary to know the mental status of psychiatric residents in uncertain situations during the COVID-19 pandemic. We aimed to observe anxiety and depression in psychiatric residents of Faculty of Medicine, Public Health and Nursing, Universitas Gadjah Mada who worked in Sardjito Hospital during the COVID-19 pandemic. This quantitative non-experimental study was conducted using cross sectional and descriptive analytic design. All 45 psychiatry residents of Universitas Gadjah Mada participated in the study and completed the Taylor Manifest Anxiety Scale (TMAS) and Patient Health Questionnaire-9 (PHQ-9). We used Chi-squared test, Student T-test and Pearson Correlation to understand correlations between demographic characteristics with depression and anxiety level. We found that only age was negatively correlated with the degree of anxiety (r =-0.364, p = 0.014). However, some other variables might be correlated with several behaviors related to anxiety and depression. More study is needed to find the reasons for these correlations and to prevent mental health problems in residents.

Background

COVID-19 has caused significant impact on the mental health of students and teaching staffs. It is likely that the most affected courses during the pandemic are the courses which involve the students to manage the health of the populace, i.e. medicine, nursing, and specialist education. Psychiatry residents in Indonesia are general physicians studying and managing patients in the hospitals to become a psychiatrist. The educational process requires the resident to manage patients with mental health problems which includes COVID-19 patients, relatives of COVID-19 patients, other patients with symptoms of dyspnea, and even other physicians.

Research data revealed that around 29.6% (24.3-35.4%) of the population were having psychological distress, 31.9% (27.5-36.7%) were having anxiety, and 33.7% were having depression. Depressive score was correlated with age, marital status, education, anxiety, sleep quality, impact of the recent events, and several post traumatic symptoms (Peng et al., 2020; Salari et al., 2020).

Health care practitioners were at higher risk of developing mental health problems. Health care practitioners who had cared for Covid-19 patients were more likely to have acute or posttraumatic stress (Odds Ratio (OR) 1.71 (1.28-2.29)) and psychological distress (OR 1.74 (1.50-2.03)) (Kisely et al., 2020). The prevalence of depression and posttraumatic stress disorder among emergency room medical workers were around 25.2% and 9.1%, respectively (Song et al., 2020). Meta-analysis by R. Pan, Zhang and J. Pan (2020) also found that the medical workers were more anxious than the general population (Standard Mean Difference = 1.145 (0.705-1.584), p=0.001). Female gender, years of working >10 years, history of mental disorder, concomitant disorder, having relatives with confirmed/suspected COVID-19, receiving support from the hospital, and hospital attempt to prevent nosocomial infection were correlated with higher levels of psychological distress (Zhu et al., 2020).

In order to optimally manage the mental health of patients, psychiatry residents need to maintain their own mental wellbeing. As part of the medical community, psychiatry residents were also exposed to news of COVID-19 pandemic and news of colleagues’ deaths due to the pandemic. Residents who live with their family and children may need to be more careful about infection prevention and use better protective gear than the ones provided by the hospitals.

There are limited research data regarding mental wellbeing of psychiatry residents during the pandemic. We aimed to measure levels of depression and anxiety of psychiatric residents and the relation with several demographic criteria related to the COVID-19 pandemic.   

Research Method

Participants and Procedures

This quantitative non-experimental study was conducted using a cross sectional and descriptive analytic design. All 45 psychiatry residents of Universitas Gadjah Mada participated in the study and completed self-reported demographic questionnaires, Taylor Manifest Anxiety Scale (TMAS) and Patient Health Questionnaire-9 (PHQ-9). The study was done on May 10^th, 2020.

TMAS is a 50 item self-report questionnaire originally developed by Taylor (1953).The sShorter version of TMAS with less items are available, however, the 50-item TMAS is the most widely used in Indonesia. The questionnaire has been adapted in the Indonesian language and tested valid and reliable with Cronbach’s alpha 0.912 (Utari, 1978).

PHQ-9 is a self-reported questionnaire developed by Kroenke and Spitzer (2002). The questionnaire has been adapted in Indonesian language and tested valid and reliable with Cronbach’s alpha 0.714 (Fatimah, 2014).

We added questions related to COVID-19, such as: “Have you managed confirmed COVID-19 patients?”, “Do you often access information about COVID-19?”, “Do you feel uncomfortable and confused about the pandemic?”, “Are you still managing any patients?”, “Do you think that the provided protective gear is enough to prevent infection?”, “Do you feel the loss from not being able to attend communal prayer?”, “Do you live with your family?” and “Are you afraid to infect your family?”

As the study was done during the time of “work-from-home” enforcement by the government, we used a video-conferencing program and an online word processor to ensure that the participants completed and understood the questionnaires from home.  

Data Analysis

Chi-squared tests were used to analyze whether or not each demographic variable affects the level of depression or anxiety. We used student’s t-test and Pearson’s correlation as an alternative if the chi-squared result was not valid.

Ethics Approval

This research was conducted in accordance with the Declaration of Helsinki. Subjects were given detailed explanation and were asked to sign informed consent forms. Subjects could withdraw from the research at any time. This research received ethical clearance from the Medical and Health Research Ethics Committee of the FKKMK UGM – Dr. Sardjito Hospital with reference number: KE/FK/0695/EC/2020.

Results

Demographic criteria of the subjects are shown in Table 1. Chi-squared analysis was used to test whether or not the distribution of depressive and anxiety levels was affected by each demographic variable. However, the results showed that in all demographic variables, 50%-83.3% of cells have expected count less than 5, and thus, chi-squared test was not valid for this population. The chi-squared analysis revealed that unmarried residents were more likely to have a more severe case of depression (x²= 9.00, p= 0.03), however, this result was not shown when the statistical analysis was repeated with different methods.

Pearson’s correlation test was used as an alternative analysis for age and student’s t-test for the rest of the variables. We found that age was negatively correlated with anxiety score and anxiety score was significantly correlated with depressive score (Table 2). The remaining categorical variables did not cause any significant difference in anxiety level nor depressive level (Table 3).

Although the number of subject is limited to survive Bonferroni correction, we performed post-hoc analysis to understand individual items related to resident anxiety and depression. The limited sample size was actually directly related to the total population of psychiatry residents in Indonesia, and thus, comprised of almost a tenth of nationwide population. We assumed that limiting the analysis because of the sample size might cause us to lose sight of important data instead.

We found that although there was no significant difference in total anxiety and depressive score, some COVID-19-related demographic variables, i.e. discomfort toward pandemic, feeling the loss of communal prayer, feeling inadequacy of protective gear, living with family, and fear of infecting family, increased the scores of one or more of the individual items, which might slightly increase the possibility of having depression or anxiety. The remaining variables caused mixed results, increasing the risk of depression in some items while being protective variables in other items.

Residents who were uncomfortable with the pandemic were more likely to feel more nervous than other people (TMAS#3, mean difference -0.17949; p=0.006), have frequent headaches (TMAS#4, mean difference -0.20513; p= 0.03), worry about financial problems (TMAS#7, mean difference 0.30769; p=0.000), worry about unpleasant possibilities (TMAS#11, mean difference 0.17949; p=0.006), have frequent worries (TMAS#27, mean difference 0.23077; p=0.002), wish to be as happy as other person (TMAS#28, mean difference 0.28205; p=0.000), be irritable (TMAS#29, mean difference -0.15385; p=0.012), easily cry (TMAS#30, mean difference 0.023; p=0.023), worry about meaningless matters (TMAS#37, mean difference 0.33333; p=0.000), feel useless (TMAS#40, mean difference 0.15385; p=0.012), be a shy person (TMAS#42, mean difference 0.17949; p=0.006), have less appetite (PHQ#5, mean difference 0.35897; p=0.000), feel less confident (PHQ#6, mean difference 0.20513; p=0.010), and have change of motoric activity due to anxiety or depression (PHQ#8, mean difference 0.15385; p=0.012).

Residents who thought that provided protective gear did not help were more likely to worry about financial problems (TMAS#7, mean difference         -0.33333; p=0.039). Residents who did not feel the loss of communal prayer were more likely to have sleep disturbance (TMAS#23, mean difference -0.17241; p=0.023) and feel increased excitation during sleep time (TMAS#35, mean difference -0.13793; p=0.043). Residents who lived with family were more likely to have frequent headaches (TMAS#4, mean difference -0.25000; p=0.003), feel tense during work (TMAS#5, mean difference 0.21875; p=0.006), feel more sensitive than others (TMAS #26, mean difference 0.31250; p=0.001), feel irritable (TMAS#29, mean difference -0.18750; p=0.012), and easily cry (TMAS#30, mean difference 0.15625; p=0.023).

The residents who were afraid to infect their family were more likely to easily cry (TMAS#30, mean difference 0.17857; p=0.022) and have changes in motoric activity due to depression or anxiety (PHQ#8, mean difference 0.17857; p=0.022).

Discussion

COVID-19 pandemic has caused significant distress to physicians and other healthcare providers, including residents. However, we found that gender, marital status, and several COVID-19 related demographic variables did not cause significant difference in total depressive nor anxiety score. We assumed that the medical community as a whole were exposed to stressors related to the COVID-19 pandemic without regards to each demographic variable. It was also important to mention that despite not meeting confirmed COVID-19 patients, residents were also managing patients with dyspnea, who might be a probable suspected case, and were also exposed to risk of infection from the medical community.

We found that the average anxiety level of psychiatry residents was relatively lower compared to healthcare providers in several other studies. The average score of TMAS was 9.02 +- 7.1, which belonged to the minimal anxiety category. There were only 3 residents (6.7%) with TMAS score 20-40 (moderate anxiety) and no residents with severe anxiety. The percentage was lower than other research which found that anxiety among healthcare provider, including residents, was 13%-70%, while the prevalence of anxiety in the general population was very heterogenous making it difficult to make any conclusions (1-82%) (Gilan et al., 2020; Li, Ye, Du, Wei, & Hei, 2020).

Average depressive score of psychiatry residents was similar to healthcare providers in other studies. Our research found that the average PHQ9 score was 2.96 +-3.12 which belonged to no depression – mild depression category. The percentage of mild-severe depression was 20% which was higher than the general population (3-20%) and similar to other healthcare providers (12-50%) (Gilan et al., 2020; Li, Ye, Du, Wei, & Hei, 2020). However, it was lower than prevalence of depression in the general population from meta-analysis data by Salari et al. (2020), which was 33.7%.

One of the possible explanations was that despite the majority of residents feeling discomfort about the COVID-19 pandemic, many had developed coping strategies to maintain a low level of anxiety. However, the coping mechanism was less effective in preventing depressive symptoms.

We also found that several COVID-19-related variables caused a change in behaviors related to anxiety and depression. Although the change was not enough to cause significant difference in the overall scores, they might reflect early symptoms of possible mental health disturbance. This study demonstrated that biological, psychological, socioeconomic, and spiritual factors were all involved in predicting responses to the pandemic.

Residents who were uncomfortable with the current pandemic situation were likely to have several behavior changes related to anxiety and depression compared to residents who had adapted to the situation. Residents who were uncomfortable with the pandemic and thought that provided protective gear was not reliable were more likely to consider financial problems. The possible explanation was that some residents were attempting to get higher level of protective gear using personal money to better protect themselves, which might cause worry about financial issues.

Residents who felt the loss of communal prayer were more likely to sleep better and be less excited during sleep time. This seemingly contradictory association could be explained if we assumed that residents who lost communal prayer were actually residents who were more religious. This could possibly mean that spiritual coping might protect against sleep disturbance. Another study also found that sleep latency was correlated with religious activities (r=0.235, p=0.044) (Khoramirad, Mousavi, Dadkhahtehrani, & Pourmazi, 2015).

Information-seeking activity did not cause any significant change in any of the items, but there was only a near-significant difference in TMAS item 26, which means that residents who searched for more COVID-19 information were possibly more sensitive than others (p=0.051). It was likely that regardless of information-seeking activity, residents were already exposed to a large number of sources of distressing information.

Our results were different from research by Peng et al. (2020), which mentioned that age, marital status and personality were correlated with depressive symptoms. It was also different from research by Zhu et al. (2020) which found correlation between family factors and mental health and research by X. Xiao, Zhu, Fu, Hu, Li and J. Xiao (2020) which found correlations between protective measures and contact history with anxiety and depression. It was possible that our population had a different set of risk factors which have not been explored before, and more study is necessary to further explore the possibilities.

Implications

Among the independent variables, only age might have the effect to protect against anxiety during the pandemic. However, several other variables might alter the response and could possibly cause mental health problem in the future if not properly managed. Residents who were younger, having discomfort about the pandemic, having problems with protective gear, having ineffective spiritual coping, living with the family, and afraid of infecting the family might require periodic assessment and mental health support when necessary.

Limitations

The study was done using self-reported questionnaires. Although it was more appropriate during the pandemic situation, this approach might be less accurate in properly diagnosing mental health problems. Better assessment tools might be necessary when the pandemic had been better-controlled. Although the study used whole-sampling, data from residents in other hospital and from another field of study might be required for comparison.

Declarations

This study does not have potential conflict of interest.

References

 Fatimah, N. (2014). Perbedaan antara obesitas dan non obesitas terhadap kejadian depresi pada ibu rumah tangga di daerah Kelurahan Cililitan Jakarta Timur. Syarif Hidayatullah State Islamic University Jakarta.

Gilan, D., Röthke, N., Blessin, M., Kunzler, A., Stoffers-Winterling, J., Müssig, M., … Lieb, K. (2020). Psychomorbidity, resilience, and exacerbating and protective factors during the SARS-CoV-2-pandemic. Deutsches Aerzteblatt Online. https://doi.org/10.3238/arztebl.2020.0625

Khoramirad, A., Mousavi, M., Dadkhahtehrani, T., & Pourmarzi, D. (2015). Relationship between sleep quality and spiritual well-being/religious activities in Muslim women with breast cancer. Journal of Religion and Health, 54(6), 2276–2285. https://doi.org/10.1007/s10943-014-9978-0

Kroenke, K., & Spitzer, R. L. (2002). The PHQ-9: A new depression diagnostic and severity measure. Psychiatric Annals, 32(9), 509–515. https://doi.org/10.3928/0048-5713-20020901-06

Li, S., Ye, Z., Du, C., Wei, Q., & He, C. (2020). The residents’ mental health status and community’s role during the COVID-19 pandemic: a community-based cross-sectional study in China. Annals of Translational Medicine, 8(20), 1321–1321. https://doi.org/10.21037/atm-20-6687

Pan, R., Zhang, L., & Pan, J. (2020). The anxiety status of Chinese medical workers during the epidemic of COVID-19: a meta-analysis. Psychiatry Investigation, 17(5), 475–480. https://doi.org/10.30773/pi.2020.0127

Peng, M., Mo, B., Liu, Y., Xu, M., Song, X., Liu, L., … Zhang, X. (2020). Prevalence, risk factors and clinical correlates of depression in quarantined population during the COVID-19 outbreak. Journal of Affective Disorders, 275(May), 119–124. https://doi.org/10.1016/j.jad.2020.06.035

Salari, N., Hosseinian-Far, A., Jalali, R., Vaisi-Raygani, A., Rasoulpoor, S., Mohammadi, M., … Khaledi-Paveh, B. (2020). Prevalence of stress, anxiety, depression among the general population during the COVID-19 pandemic: a systematic review and meta-analysis. Globalization and Health, 16(1), 57. https://doi.org/10.1186/s12992-020-00589-w

Song, X., Fu, W., Liu, X., Luo, Z., Wang, R., Zhou, N., … Lv, C. (2020). Mental health status of medical staff in emergency departments during the Coronavirus disease 2019 epidemic in China. Brain, Behavior, and Immunity, 88(January), 60–65. https://doi.org/10.1016/j.bbi.2020.06.002

Taylor, J. A. (1953). A personality scale of manifest anxiety. The Journal of Abnormal and Social Psychology, 48(2), 285–290. https://doi.org/10.1037/h0056264

Teng, Z., Wei, Z., Qiu, Y., Tan, Y., Chen, J., Tang, H., … Huang, J. (2020). Psychological status and fatigue of frontline staff two months after the COVID-19 pandemic outbreak in China: A cross-sectional study. Journal of Affective Disorders, 275(May), 247–252. https://doi.org/10.1016/j.jad.2020.06.032

Utari, D. (1978). Taylor Manifest Anxiety Scale. University of Muhammadiyah Malang.

Xiao, X., Zhu, X., Fu, S., Hu, Y., Li, X., & Xiao, J. (2020). Psychological impact of healthcare workers in China during COVID-19 pneumonia epidemic: a multi-center cross-sectional survey investigation. Journal of Affective Disorders, 274(May), 405–410. https://doi.org/10.1016/j.jad.2020.05.081

Zhu, Z., Xu, S., Wang, H., Liu, Z., Wu, J., Li, G., … Wang, W. (2020). COVID-19 in Wuhan: Sociodemographic characteristics and hospital support measures associated with the immediate psychological impact on healthcare workers. EClinicalMedicine, 24, 100443. https://doi.org/10.1016/j.eclinm.2020.100443

Tables

Table 1

Demographic Characteristics

Source: Primary Data  

Table 2

Pearson’s Correlation of Age, Anxiety Score, and Depressive Score

* denotes significant correlation with p<0.05

** denotes significant correlation with p<0.01 

Table 3

Student’s T-test Result to Compare Anxiety and Depressive Score in Each Subgroup

Source: Primary Data. No significant difference was found in all variables