A Cross-Sectional Questionnaire-Based Study to Assess Anxiety among Healthcare Professionals in India during COVID-19 Pandemic

Abstract

Background and Aims: 

Coronavirus has resulted in unique challenges to the healthcare professionals. The pandemic has affected the mental health of healthcare professionals, which may impact their ability to provide care to patients. We aimed to assess anxiety among healthcare professionals, and its relation to selected parameters.

Methods: 

This online questionnaire-based study recruited Indian healthcare workers through convenience sampling. Anxiety was measured using the Generalized Anxiety Disorder 7 scale. The reasons of emotional distress were also ascertained.

Results: 

The sample comprised of 677 healthcare professionals including resident doctors, consultants, interns, nursing personnel, and others. Moderate/severe anxiety was present in 24.4% of the respondents. Moderate/severe anxiety was more likely to occur among those who had a friend or family member affected by COVID-19 (odds ratio [OR] 2.74, 95% confidence intervals [CI] 1.61–4.69), or who themselves had a psychiatric disorder in the past (OR 2.44, 95% CI 1.36–4.38). Those who reported long working hours and inadequate sleep, fear of being infected by the virus, concerns about transmitting the virus to a family member, mob violence against doctors, and shortage of personal protective equipment had significantly higher odds of moderate/severe anxiety.

Conclusion: 

A considerable proportion of healthcare professionals suffering from anxiety suggests the need for addressing their mental health concerns.

Introduction

As the number of COVID-19 cases soar worldwide,^[1] the effects of the pandemic have been seen in various domains of life, including healthcare services. Health care workers comprise an important vulnerable group of individuals, as the potential overwhelms of our healthcare systems need them for longer hours in strenuous circumstances and making difficult decisions.^[2] They have been grappling with new and unforeseen challenges like shortage of personal protective equipment and an overload of information from unvetted sources, in addition to the challenges of an inadequate infrastructure to deal with the rapidly escalating number of cases. Apart from this, the risk of infection and transmitting the virus to family members is a burden that they have to constantly deal with. This is likely to lead to psychosocial distress, which can lead to consequences such as impaired work performance and poor patient outcomes.

It is crucial to ascertain the levels of anxiety in the population engaged in providing health care since they are the forerunners of the battle against COVID-19. Studies have started to emerge which have reported on the rates of anxiety among healthcare professionals in this period of the pandemic. It has been estimated that anxiety may be prevalent in up to one-fourth of healthcare professionals.^[3] Several studies from India and the world have assessed anxiety and other symptoms of psychological distress in Indian healthcare professionals and healthcare professionals from other parts of the world.^[4–9] This study was conducted in the initial phase of the pandemic (i.e., April to July 2020) to assess anxiety and predictors of such anxiety among health care professionals in India.

Methodology

The present cross-sectional observational study used an online questionnaire. The inclusion criteria were health care professionals (including doctors, nurses, and paramedical staff) working in India. Exclusion criteria were failure to provide consent or provided <50% of the responses. The recruitment of the participants was through convenience online sampling using social media and informed consent was obtained through the Google Forms-based questionnaire. The survey was voluntary and no monetary or nonmonetary incentives were offered for participation. The study received institutional ethics committee approval.

The questionnaire comprised of questions pertaining to age, gender, current work location, work profile, whether there have been cases/deaths related to COVID-19 in their hospital, whether they were taking care of COVID-19 patients or likely to do so in the future, and whether their friends or family members contracted COVID-19 or died/went into critical condition due to it. Respondents were also asked whether they have ever suffered from any diagnosed psychiatric illness (as per self-report), and if so, whether their condition worsened, remained the same, or improved in the recent past. Anxiety was assessed using the English version of the Generalized Anxiety Disorder 7-item (GAD-7) scale which is a standardized and validated brief 7 item scale to screen for anxiety. Total scores range from 0 to 21, and higher scores reflect a greater extent of anxiety. A cut-off of 10 is generally considered as indicating moderate to severe anxiety.^[10] Respondents were also asked about their reasons of emotional distress using options developed by the investigators based upon current common distressing issues. The respondents were given the opportunity to add more reasons if they found relevant. There were a total of 21 questions over two screens and the questions were nonrandomized. Appropriate nonresponse answer choices like “Don’t know,” “Prefer to not answer,” and “Not applicable” were included to encourage completion. Compulsory fields or adaptive questions were not used.

Efforts were made to distribute/advertise the survey to a nationally representative population using Indian Medical Association groups, resident forums, healthcare staff groups, medical college alumni groups on Whatsapp, Facebook, and Instagram. Participants were recruited through convenience sampling. Responses were recorded from over 13 states and Union Territories but a majority of the responses came from Delhi, Punjab, Gujarat, Tamil Nadu, and Maharashtra. The survey was opened on April 20, 2020, and closed on July 4, 2020. A formal sample size estimation was not done, though we had aimed for a sample of >500. CHERRIES guidelines^[11] were used for reporting of data. No identifying information was collected and the data access was restricted only to the investigators in the form of a private Google sheets document. Before circulation of the form, it was tested on computers, tablets, and mobile devices to ensure smooth functioning. The survey was open (not password-protected). The uniqueness of visitors was ensured using E-mail account logins (E-mail addresses were not collected but anonymous account logins prevented multiple responses by a single participant). As an in-built feature, Google Forms offered a review step to modify the answers before submission of the form. We did not use cookies or check IP addresses, and information about view rate, participation rate, and the completion rate is not available due to the nature and platform of the survey. Participants with <50% of the responses were excluded. Logfile analysis was not applicable and we did not find any atypical timestamp, while no statistical correction was applied.

The analysis was done using SPSS version 21 (IBM Corp, Armonk, NY, USA). Descriptive data were represented by mean, standard deviation, frequencies, and percentages. Since the GAD 7 scores were not normally distributed, the median and inter-quartile range was used to represent the data. The sample was divided into those with moderate/severe anxiety (GAD 7 scores of 10 or more) or those without moderate/severe anxiety (GAD 7 scores of <10). Binomial regression analysis was used to find the unadjusted and adjusted odds ratios of anxiety across various parameters. Missing value imputation was not conducted, and a P < 0.05 was considered significant.

Ethical approval

The study was accorded Ethical Committee Approval vide Ethics Committee IEC-311/27.04.2020, RP-16/2020 dated 15.05.2020. Informed consent was taken from all the participants. The study was carried out in accordance with the principles as enunciated in the Declaration of Helsinki.

Results

Out of the 724 responses, we excluded 47 (6.4%) as they did not pertain to healthcare workers working in India. The characteristics of the remaining 677 respondents are presented in Table 1. The median age was 28 years. Half of the respondents were male and resident doctors comprised the largest group of the respondents. Less than a tenth of the participants reported to have a psychiatric illness in the past (n = 58, 8.6%), with depression being the most common diagnosis (n = 33, 4.9%). Of those who had a psychiatric illness ever, about a fourth reported a worsening in the recent past.

Table 1: Characteristics of the sample (n=677)

Normality assumption was not met for GAD7 (one sample Kolmogorov–Smirnov test P < 0.001). The median GAD7 score was 5 with an interquartile range of 2–9. Moderate/severe anxiety reflected by scores of 10 and above in GAD7 was present in 165 participants (24.4% of the respondents). In this group of 165 participants, doing work, taking care of things at home, and getting along with others was found to be not difficult at all by 6, somewhat difficult by 77 and very difficult by 82 participants (3.6%, 46.7%, and 49.7% respectively). Moderate/severe anxiety was found in 23.6% of residents, 19.0% consultants, 33.1% interns, 21.8% nurses, and 21.6% of other respondents.

The relationship of having moderate/severe anxiety with various others parameters is presented in Table 2. In unadjusted bivariate relationship, it was seen that younger age, being an intern, having COVID-19 cases in the hospital, having a friend or family member either diagnosed with COVID-19 or being critically ill/dead due to COVID-19, or having a diagnosed psychiatric disorder in the past was associated with having moderate/severe anxiety (i.e., GAD7 scores >10). On adjusted analysis, only having a friend or family member diagnosed with COVID-19 or having a diagnosed psychiatric disorder themselves were associated with having moderate/severe anxiety.

Table 2: Relationship of anxiety with other factors

The participants were asked about the reasons of emotional distress to them. The responses from the structured options are presented in Table 3. The most commonly reported reasons of emotional distress were concerns about transmitting the virus to a family member and fear of a loved one being infected with the virus. These were also the most important reasons of emotional distress among the respondents. Among the reasons, long working hours and inadequate sleep, fear of being infected by the virus, concerns about transmitting the virus to a family member, mob violence against doctors, and shortage of personal protective equipment were associated with significantly higher odds of having moderate/severe anxiety.

Table 3: Reasons of emotional distress (n=677)

Discussion

One of the major findings of this brief report has been that moderate/severe anxiety was present in about a fourth of the population of healthcare professionals. The prevalence rates of anxiety have been reported to range from 10.5% to 44.5% among healthcare workers according to the review by Pappa et al.^[3] This review included only studies from China and subsequent studies have emerged which have reported rates of anxiety in the range as presented above.^[12–14] There are several studies from India as well which have assessed the extent of anxiety in healthcare professionals. While Chatterjee et al. report anxiety in 39.5% of the population,^[4] Sharma et al. suggested anxiety to be present in about 56.7% of the healthcare workers,^[5] and Suryavanshi et al.^[7] reported anxiety to be present in about 50% of the healthcare professionals. These studies might have found higher rates of anxiety probably due to differences in the period of the pandemic in which data collection occurred, and possible selection biases. On the other hand Wilson et al.^[8] have reported lower rates of anxiety in the studied healthcare provider population.

One of the predictors of having moderate/severe anxiety in this population was having a diagnosis of psychiatric illness ever in their lifetime. This suggests those who have had mental health issues in the past, were likely to appraise the current situation with greater anxiety, or their previous disorder might have flared up, which manifested as anxiety. Only a fourth of the individuals with a psychiatric diagnosis considered their condition to have deteriorated, which may again signal a relapse of their previous condition. Yet, those individuals who have suffered through a psychiatric illness emerge as a vulnerable group and need attention and appropriate help to deal with the emergence of anxiety. Alternately, it is also possible that those who have had a psychiatric illness in the past had lesser response bias, and answered the questions much more freely. Having coronavirus in the family members has been reported as an important cause of anxiety in the present sample. Fear of passing the virus to the family members and fear of a loved one being infected by the virus were important reasons of emotional distress in the participants. Similar findings from China suggest that a majority of the participants were worried about family members contracting COVID-19.^[15,16]

The findings of the present study highlight that moderate/severe anxiety has been quite prevalent among healthcare professionals in India. This may not only impair their quality of life and lead to distress and dysfunction, but may also result in subtle or overt impairments in ability to care for the patients. Measures to address anxiety through online consultations, psychotherapy, adequate off-time, breaks during shifts to be in touch with family members, and psychotropic medication when required might help in the reduction of anxiety. Also, those who have had psychiatric illnesses may like to have a closer follow-up for their mental health condition.

Some limitations of the study should be considered while drawing inferences from the findings. The sampling through social media may have resulted in biases in sample selection, and we do not have information of response rate. The choices given as reasons of emotional distress did not derive from a prevalidated questionnaire, and the GAD7 cut-offs have been used from a different population. There was an over-representation of doctors, and the results may not translate into all categories of medical professionals. The study was cross-sectional in nature and hence cannot capture the dynamic nature of changing mental states with the evolution of pandemic. Sample size was small. The study was conducted through an online survey, and not through face-to-face interview. The nature of the assessment was brief and detailed assessment was not carried out.

To conclude, the present study suggests that during the pandemic, one in four healthcare professionals in India had moderate/severe anxiety symptoms. Future work should look at the trajectory of anxiety as the pandemic progresses, what measures are accepted and employed by healthcare professionals to reduce their anxiety, and how anxiety interacts with other related constructs like depression and stress.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.