Discussion
The main results of the present study indicated a high prevalence of death in individuals with SARS due to COVID‑19 who required IMV. The mortality rate was higher in individuals aged >60 years who received azithromycin during SARS-CoV‑2 infection.
The prevalence of death (46.1%) in individuals hospitalized with SARS due to COVID‑19 was similar to that reported in other studies2,8. Notably, the patients included in the present study were critically ill and treated with IMV on admission to the ICU. These data are concerning and underscore that the high mortality due to COVID-19 in critically ill patients may be because of the continuous need for respiratory support and long stays in the ICU2.
The findings of the present study confirm that the survival of critically ill patients with COVID-19 is particularly low in elderly men. Moreover, age of >60 years was associated with in-hospital death. A systematic review also stated that advanced age has been recognized as an important risk factor for COVID‑19 mortality11. One explanation is that aging leads to impaired functioning of multiple body systems, including the immune system, which is a factor involved in the increased mortality due to COVID-19 in the elderly11.
Although mortality was higher in men, no statistical significance was observed in the present study. Grasselli et al. found a relationship between the male sex and increased mortality in patients with COVID‑19 admitted to ICUs2. In addition to higher mortality, another large study involving 3.1 millions of patients with COVID-19 reported that the male sex was associated with a higher ICU admission rate12. This phenomenon is attributed to the fact that women have a higher number of CD4+ T cells, a more robust cytotoxic activity of CD8+ T cells, and greater production of immunoglobulin by B cells than men, enabling them to produce a more efficient cellular and humoral response12.
The use of azithromycin was also associated with in-hospital death. Azithromycin is an antibiotic with anti-inflammatory and antiviral properties, and it was thought to have activity against SARS-CoV‑213. However, based on different studies, a systematic review reported that azithromycin, along with other medications, including angiotensin-converting enzyme inhibitors, aspirin, colchicine, hydroxychloroquine, inhaled corticosteroids, intranasal corticosteroids, interferon beta, ivermectin, lopinavir-ritonavir, and vitamin C, has no important benefit on any important outcome for patients with COVID‑1914. Based on the present study design, this association between azithromycin and death does not indicate causality.
Dexamethasone may be beneficial in patients with COVID‑19, mainly in more severe forms with exacerbated inflammatory activity, because of its potential anti-inflammatory effect, which confers it the ability to decrease gene transcription of several proinflammatory cytokines, chemokines, and adhesion molecules by inhibiting the generation and release of these mediators. However, dexamethasone may also prevent B‑cell-mediated antibody production and reduce T‑cell immune function, which may result in a higher plasma viral load and an increased risk of secondary infections15. Although the present study did not report an association of dexamethasone use with mortality, a randomized clinical trial in patients hospitalized with COVID‑19 in the United Kingdom comparing the use of 6 mg dexamethasone once daily for 10 days to placebo reported reduced 28-day mortality rate in patients with COVID-19 receiving either IMV or oxygen therapy without IMV, with no impact on mild cases without respiratory support16.
Based on the current evidence, low-dose systemic steroids may be considered for specific patients with COVID-19 who are critically ill or require supplemental oxygen. However, routine use of corticosteroids should be avoided, particularly in patients with mild symptoms or in the early stages of the disease unless indicated for another reason, such as those related to an individual’s condition17,18. The retrospective nature of this study and lack of standardization of dose and duration of therapy may have contributed to the absence of any association of corticosteroid use with mortality in the present study. This is because any use of dexamethasone, whether during the ICU stay or earlier, was considered in this study.
The prone position contributes positively to the ventilation-perfusion ratio and to the recruitment of dependent lung segments, culminating in the opening of collapsed dependent alveoli and thereby providing better gas exchange and oxygenation. Among mechanically ventilated non-COVID-19 patients with severe acute respiratory distress syndrome, those who were ventilated in the prone position had a lower mortality rate.19 The prone position can reduce the relative fraction of the pulmonary shunt by 30% compared with the supine group in patients with injured lungs20. A recent meta-analysis demonstrated that the prone position improved the PaO₂/FiO₂ ratio, with better SpO₂ than the supine position, in patients with COVID‑1921. Despite the aforementioned beneficial effects, the present study found an association between pronation and death, which may actually reflect the greater severity in patients who were pronated to improve ventilation and gas exchange. However, after the multiple regression analysis, pronation was not confirmed as an independent factor associated with mortality.
The limitations of this study include the observational cross-sectional design, which made it impossible to establish a causal relationship between the variables. In addition, the data collection considered a sample from a medium-sized city in the State of Goiás; therefore, generalization of the results to the rest of the Brazilian population must be done with caution. Variables that included previous lung disease had a low prevalence; therefore, it was impossible to establish an association with the outcome variable or to associate the use of drugs, such as chloroquine and hydroxychloroquine, with mortality. In addition, variables, such as the length of hospital stay, ventilation parameters, and laboratory tests, were not examined. The authors recommend examining these variables in future studies.