This study is the first to track the within-person changes in the levels of anxiety in healthcare workers during two time points of the COVID-19 pandemic. This study was based on an Argentinean sample of healthcare workers, to analyze the changes in the anxiety outcomes, adjusting for main demographic factors, region, and some relevant health-related factors.
Consistent with our first hypothesis, we found higher levels of anxiety among healthcare workers from the metropolitan region compared to those from the region inside the country at the first time point of this study, when there were more than five thousand of COVID-19 cases in the former region and almost no cases in the latter region. Nonetheless, these differences disappeared in the follow-up, when the pandemic has progressed. By then, the metropolitan region reached 417,677 COVID-19 cases and the region inside the country reached 15,670 cases17. This meant around 2610 and 2328 COVID-19 cases per 100,000 inhabitants in the former and the latter regions respectively. Thus, such numbers of COVID-19 cases in both regions implies a worrying worsening of the health situation and a pressing workload for healthcare workers, which could result in higher burden of negative mental health outcomes such as anxiety. Indeed, as it was expected according to our second hypothesis, we found increasing levels of anxiety in healthcare workers from the first time point to the follow-up. Overall, these findings suggest that irrespective of the starting point in anxiety levels among healthcare workers, an increasing anxiety outcome may be expected among them as the pandemic progresses.
Furthermore, beyond the negative impact that implies for healthcare workers an increasing curve of the pandemic itself (e.g., overwhelming workload, working under extreme pressures, depletion of personal protective equipment, covering additional shifts, among others)8,18, our findings warn that the increase of anxiety among healthcare workers is mediated by the concern regarding the COVID-19 contagion. Both negative mental health outcomes (i.e., anxiety) and relevant predictors for such outcomes (i.e., COVID-19 contagion) had been suspected to occur during the current pandemic based on previous epidemic and disease outbreaks19 and also based on current but cross-sectional evidence4,5,20, though they had not been supported with longitudinal evidence during the COVID-19 pandemic, until now. Published peer reviewed longitudinal studies on mental health outcomes in healthcare workers during the COVID-19 pandemic are lacking and our study is a first step in order to fill this gap.
As it was suggested by the review of Carmassi et al.19, based on previous cross-sectional and some longitudinal evidence from the Severe Acute Respiratory Syndrome (SARS) and Middle East Respiratory Syndrome (MERS) outbreaks, it may be expected to find the worst mental health outcomes among healthcare workers who had gotten sick with the COVID-19. Nonetheless, based on our current longitudinal study, the largest increases in anxiety levels are not among healthcare workers who have been gotten sick, but in those who wondered if they had been infected with the COVID-19, while being symptomatic. Considering that majority of the healthcare workers in our study were involved in direct patient care, these findings suggest that the frequent exposure to the COVID-19 in their working environment contributes to the uncertainty and suspicion about the contagion, which becomes a heavier factor for the increase of anxiety among them, than the actual contagion. Similar hints have emerged with general discomfort and psychological distress outcomes among healthcare workers during the current pandemic21. During the SARS outbreak, the perception of personal danger, for instance related to shortages of hospital masks, was exacerbated by the uncertainty in healthcare workers2. Moreover, uncertainty, although not referring to the COVID-19 contagion, has been demonstrated as related to negative mental health outcomes, namely anxiety and depression, among the general population both prior22 and during23 the COVID-19 pandemic. Leading with uncertainty is a routine part of the medical practice24 and, in a more general sense, is part of the work in all healthcare-related professions. However, this known uncertainty is usually referring to aspects pertaining to the patient (e.g., her/his diagnosis, treatment, etc.) or to the disease (e.g., its etiology, prognosis, etc.), but not to the health state of the healthcare workers themselves, like in the unprecedented COVID-19 pandemic. This last kind of uncertainty, commonly unknown to most healthcare workers, often emerges during contagious disease outbreaks. In this sense, during viral epidemic outbreaks, a higher prevalence of a number of mental health problems, mainly anxiety and depression, was associated with a higher perception of threat and risk among healthcare workers25. Fortunately, the uncertainty regarding COVID-19 contagion in healthcare workers is both a preventable and a modifiable factor by means of providing adequate protection supplies and more COVID-19 tests for them.
Furthermore, the anxiety levels of healthcare workers were mediated by the interaction effect between the region and the COVID-19 contagion. We believe that additional variables not assessed in our study may help to interpret this finding, for instance, shortages in personal protective equipment, which is a major source of distress in healthcare workers during the COVID-19 pandemic26. Such shortages are known to be related to economic reasons and hurdles in transportation and distribution logistics, among other reasons. Although the shortage of personal protective equipment affected countries all around the world during this pandemic, this problem is two-fold for regions with lower income and with more serious distribution problems27. In this regard, compared to the metropolitan region, the region inside the country is of a lower income and has remote access, since it is located in the north westernmost of Argentina. Thus, shortages in personal protective equipment may have affected healthcare institutions in this region – as it has been reported in the media (see e.g.,28) – more than institutions from the metropolitan region and such scarcity may have acted as an anxiogenic factor. Indeed, our results show that, irrespective of the time point, anxiety was the highest in healthcare workers from the region inside the country who wondered if they had been infected with the COVID-19, whether they were asymptomatic or symptomatic.
Finally, we have also found an interaction effect between the mental disorder history and the COVID-19 contagion on anxiety. Strikingly, the highest anxiety affected healthcare workers without mental disorder history who wondered if they had been infected with the COVID-19 and were symptomatic. This is contrary to what would be expected based on studies during the SARS outbreak, which highlighted a relationship between having a history of psychiatric disorders and worst mental health outcomes among healthcare workers29,30. However, such outcomes would be, in turn, inversely associated with years of health care experience and the perceived adequacy of training and support29, variables that we have failed to assess in our study and may help to explain the opposite findings. Nonetheless, the cited studies are not entirely comparable to our study. For instance, they were based on smaller samples than ours, which were comprised of either 102 nurses solely30 or 139 healthcare workers29. In addition, the findings of the latter study correspond to a follow-up of one to two years after the SARS outbreak, rather than a follow-up during the sanitary event as in our study. All in all, the interaction effect between the mental disorder history and the COVID-19 contagion emerged in our study would suggest that the anxiety outcomes found in healthcare workers are mainly due to the concern about the COVID-19 contagion, rather than due to pre-existing mental health vulnerabilities.
Limitations
The findings of this study, although valuable, should be considered in the context of some limitations. First, as we discussed earlier in this paper, additional factors not measured in this study may be relevant to thoroughly understand the negative mental health impacts in healthcare workers during the COVID-19 pandemic and further research is needed to address such factors. Second, the sample was not representative of all Argentinean healthcare workers, though it included data of healthcare workers from well-balanced metropolitan and non-metropolitan or rural areas and there was a low attrition between the two measurements. Third, most women participated. Although to some extent this reflects the fact that women are majoritarian among Argentinean healthcare workers31, the uneven sex distribution of the sample is a potential bias source. Fourth, the instrument that we used to measure anxiety, though is a validated screening tool, is not a clinical assessment tool. Fifth, we failed to measure anxiety prior to the COVID-19 pandemic; thus, we cannot assure that the outcome of interest was not present at the start of the study. However, the longitudinal design we used allowed us to ascertain that such outcome meaningfully increased as pandemic progresses, irrespective of the different starting anxiety level of healthcare workers from both regions.
Implications
Despite the above-mentioned limitations, this study provides valuable longitudinal evidence-based knowledge on the within-person changes in the levels of anxiety of healthcare workers during the COVID-19 pandemic and identifies foremost factors conducive to anxiety outcomes. The uncertainty regarding the COVID-19 contagion, a preventable and modifiable factor interacting to produce the worst anxiety outcomes among healthcare workers, should be promptly addressed by public health officials and government officials.