COVID-19 pandemic affected health professionals in many ways. The stress of the pandemic has physical and psychological health consequences, in addition to economic crisis and high unemployment rate.1,2 Different countries have faced the pandemic in different ways, and emotional consequences are under intense evaluation to assess its impact on quality of life.3

Although the main focus of frontline healthcare workers (HCW) is on minimizing transmission and treating COVID-19 patients, the pandemic impact on mental health and its consequences cannot be underestimated.4 Healthcare workers are facing increased mental health problems, such as insomnia, anxiety, depression,5,6 burnout syndrome (BS),7 and poor health-related quality of life (HRQOL).3 In Italy, frontline HCW presented with many symptoms of work-related psychological stress and burnout during the peak of the COVID-19 pandemic. Many of these professionals may be at risk of post-traumatic stress disorders.7

Some aspects, including fear of infecting relatives, long and direct care for COVID-19 patients, and sleep disorders may contribute to the risk of developing psychological symptoms.6,8 Yet, there are few studies linking anxiety to HRQOL and BS exclusively among physicians.4 Therefore, this study aimed to describe HRQOL and burnout syndrome in frontline physicians with anxiety during the COVID-19 pandemic.

This cross-sectional study was conducted from September through October 2020, during the first-wave phase of COVID-19. Questionnaires were sent electronically to 450 physicians from all COVID-19 referral institutions in the State of Bahia, Brazil using an Internet-based survey site (https://pt.surveymonkey.com/). Of these, 251 physicians answered the questionnaire. Before assessing the questionnaire, participants signed an informed consent form and agreed to participate in the survey. Twenty-eight questionnaires with more than 20% of missing data were discarded. The final sample included 223 physicians, representing 49.6% of physicians in the frontline in the study period. Sociodemographic data (sex, income, marital status, religion, and years of experience), COVID-19 related aspects, quality of like9 (WHOQOL BREF), burnout10 (Maslach Burnout Inventory, MBI) and anxiety11 (Beck Anxiety Inventory, BAI) were assessed. The research was approved by the National Ethics Review Committee of the National Council of Health, Brazilian Ministry of Health, protocol number (4.008.150).

In this sample, the clinical prevalence of anxiety (BAI≥16) in frontline physicians was 17.0% (Table 1). Anxiety was over two-fold more prevalent in female physicians than in males (PR=2.2; p = 0.007). Marital status, religion, and/or having less than two years work experience were not associated with anxiety (Table 2).

Among physicians in the anxiety group, alcohol ingestion was 87% higher (PR=1.87; p = 0.030) and use stimulants was over three-fold more prevalent than in physicians without anxiety (PR=3.33; p = 0.002). Those with anxiety were more likely to avoid close contact with patients suspected of COVID-19 (PR=1.93; p =  0.029) (Table 3).

All physicians on the frontline against COVID-19 experienced at least moderate symptoms of Burnout Syndrome (Table 4). Physicians in the anxiety group showed higher scores in emotional exhaustion (EE) (p = 0.0001) and depersonalization (DP) (p = 0.001) domains, in addition to lower scores in personal accomplishment (PA) (p = 0.004), than the group without anxiety.

All HRQOL scores in the anxiety group were lower than those without anxiety (p < 0.01). (Table 5). Scores of WHOQOL-BREF were significantly correlated with all domains of MBI in physicians without anxiety (p = 0.001). A significantly positive correlation was detected between all domains of WHOQOL-BREF and PA MBI domain in physicians with anxiety, ranging from 0.38 to 0.42 (Table 6). The Fisher r-to-z test showed differences according to anxiety in correlations between EE and Psychological Domain (z = 1.92, p = 0.03) and between DP and Psychological Domain (z = 1.76, p = 0.04).

To the best of our knowledge this is the first study that investigated the impact of anxiety on health-related quality of life and symptoms of burnout in frontline physicians in Bahia during the COVID-19 pandemic. The response rate of the present study was similar to the 54% mean of other published studies in the literature that conducted email survey with medical professionals.19 Even though the prevalence of anxiety in our sample was 17.0%, both groups experienced moderate levels of BS symptoms. Prior to the COVID-19 pandemic, a study involving intensive care physicians in Brazil showed a high prevalence of BS symptoms in EE domain, and a low prevalence in other domains of MBI.20 However, that study included a smaller sample of Brazilian physicians (24 from Belem, 18 from Goiania, 28 from Porto Alegre, 65 from Salvador and 45 from Sao Paulo). Another Brazilian study evaluated the effect of supportive leadership on symptoms of burnout in physicians during COVID-19 pandemic, showing a weaker effect for frontline physicians compared to previous data.21 A cohort study in the Netherlands, with physicians and nurses working in a critical care scenario during COVID-19 pandemic, reported a two-fold higher prevalence of burnout symptoms in comparison to anxiety.22

During the COVID-19 pandemic, nurses and physicians with increased scores for exhaustion, depression, anxiety, and stress were associated with a greater fear of being infected with coronavirus.23 Higher score of emotional exhaustion and levels of stress were seen in nurses with contact with COVID-19 patients, compared to nurses with no contact with such patients.23 In Italy, the burnout syndrome was associated with more depression or higher anxiety scores in HCW during the acute phase of COVID-19.24

In Pakistan, a study described a prevalence of 43% of anxiety/depression in frontline physicians one month after the first reported COVID-19 case. The lack of knowledge about COVID-19 was associated with depression in frontline doctors.25 Almost half (47%) of the Indian healthcare professionals, most of them directly involved in care of COVID-19 patients, had symptoms of depression, which was classified as moderate to severe in 22% of them.3 Symptoms of depression, anxiety, insomnia, and distress were reported by physicians and nurses in China two months after the first diagnosed case.26 Furthermore, burnout, anxiety, and depression symptoms had a negative effect on the quality of life of HCW during the COVID-19 pandemic.27

Taking care of COVID-19 patients may lead to a higher prevalence of anxiety, depression, and lower HRQOL28 in frontline healthcare workers. Additionally, moderate to severe depression and anxiety symptoms had a negative impact on their quality of life.3 In our study, all physicians had lower HRQOL in comparison with data found in the heatlhy Brazilian population.13 Physicians with anxiety had lower HRQOL when compared with the group without anxiety. In addition, smoking and drinking habits may be associated with poor HRQOL and poor mental health.28 Our findings indicate that alcohol ingestion, as well as use of stimulants, were higher in the anxiety group during COVID-19 pandemic.

We detected a significant correlation between WHOQOL-BREF and MBI in physicians without anxiety. Moreover, there were differences in the correlations between EE and Psychological Domain and between DP and Psychological Domain, according to anxiety. A previous study found that individuals who were predisposed to anxiety and oversensitivity to bodily sensations presented poorer levels of perceived health than those without anxiety. In the same study, not only somatic symptoms, but also symptoms of anxiety are considered a relevant target for improving HRQOL.29 The correlation between burnout and health-related quality of life in individuals with anxiety is still not well explored in literature. Anxiety disorders and its effect in HRQoL and burnout should be further investigated.

Our data show that anxiety was more prevalent among females. A systematic review reported a higher risk of female HCW in developing depression, anxiety, and insomnia compared to male workers.5 Moreover, female frontline workers reported experiencing more severe symptoms and higher levels of depression, anxiety, insomnia, and distress than their male counterparts, during COVID-19 pandemic in China.26

This study had some limitations. The cross-sectional design does not allow to confirm causality. The prevalence ratio was calculated to better evaluate our results and all applied instruments presented good reliability and showed satisfactory internal reliability evaluated by Cronbach's alpha test, according to previous psychometric evaluations of MBI and WHOQOL BREF by developers.10,12 The consistency of final results indicates that the detected associations are robust enough and demonstrates a clear association between frontline working on COVID-19 care and such emotional disturbances.

Physicians with anxiety showed increased emotional exhaustion, decreased personal accomplishment and a low health-related quality of life in WHOQOL BREF domains during the COVID-19 pandemic in Brazil. Physicians´ anxiety was associated with lower health-related symptoms. All domains of WHOQOL BREF were correlated with all MBI domains among physicians without anxiety. Differences in correlation according to anxiety were remarkable in Psychological HOQOL BREF Domain and Emotional Exhaustion and Depersonalization MBI domains. The effect of anxiety leading to poorer levels of perceived health needs to be further investigated.