Coronavirus disease 2019 (COVID-19), an infection caused by the severe acute respiratory syndrome Coronavirus 2 (SARS-CoV-2), has a wide range of clinical presentations, from asymptomatic infection to critical and fatal disease.1 Most symptomatic infections are mild and present with fever, cough, and fatigue.2

Changes in blood gas concentrations can cause dyspnea and respiratory distress, which are crucial indicators of severe disease.3 According to the Centers for Disease Control and Prevention guidelines, patients with hypoxemia are classified as having severe disease, with the requirement for more advanced medical care.4 Patients may have hypoxemia without dyspnea, also known as silent, non-dyspneic, or “happy” hypoxemia.5 In patients presenting to the hospital with COVID-19 and hypoxemia, the prevalence of this unusual clinical presentation ranges from 4.9% to 31.9%. Given the definition of silent hypoxemia, the prevalence of this severe manifestation of COVID-19 may be underestimated.6,7

Despite its prevalence and potential for rapid clinical deterioration, the pathological mechanisms and prognostic implications of silent hypoxemia remain poorly understood. This underscores the critical need for epidemiological investigations to elucidate its prevalence, risk factors, and impact on patient outcomes. Previous studies assessing the impact of silent hypoxemia in COVID-19 were based on relatively small sample sizes and may have been underpowered to detect significant risk factors associated with unfavorable clinical outcomes.6–11

In this context, we aimed to describe the clinical characteristics and predictors of mortality in a cohort of patients who were hospitalized for severe acute respiratory syndrome (SARS) due to COVID-19 presenting with silent hypoxemia.

This retrospective cohort study evaluated the clinical characteristics and predictors of death in hospitalized patients due to SARS from COVID-19 who presented with silent hypoxemia at admission. Our main findings were that (I) advanced age, presence of comorbidities, and development of respiratory distress were independent risk factors for mortality, while (II) development of cough and need for non-invasive respiratory support were independent protective factors against mortality.

Several mechanisms have been proposed to explain silent hypoxemia in COVID-19. Gattinoni et al. proposed two primary “phenotypes”: types L and H. Type L, characterized by low elastance and reduced ventilation-to-perfusion (V/Q) ratio, results from viral-induced local interstitial edema and vasoplegia.13–15 Patients with the type L lack dyspnea due to near-normal compliance.16–19 Transition to the type H, driven by heightened COVID-19 severity and inflammation, may induce dyspnea through patient self-inflicted lung injury (P-SILI).13,20 This mirrors the conditions observed in acute respiratory distress syndrome.16,21,22

In COVID-19, the absence of dyspnea may be attributed to direct viral damage to the carotid bodies and solitary tract, as the glomus cells of the carotid body exhibit high expression levels of the SARS-CoV-2 receptor, angiotensin-converting enzyme-2.7,18,19,23–26 Acute vascular distress syndrome (AVDS), resulting from endothelial dysfunction, microthrombi formation in the pulmonary vasculature, and vascular proliferation in the lungs, causes an elevated pulmonary vascular flow, leading to a significant intrapulmonary right-to-left shunt.27,28 Additionally, variability of the hypoxic ventilatory response across different populations not only explains the high prevalence of silent hypoxemia but also suggests that this phenomenon could be present in other causes of hypoxia, such as influenza pneumonia, among others (Fig. 2).16

Figure 2.

Proposed mechanisms of silent hypoxia in COVID-19.

(A) Direct viral infection of the carotid bodies and central nervous system may reduce the perception of dyspnea mediated by chemoreceptors. (B) Endothelial cell dysfunction can precipitate acute vascular distress syndrome. This fosters the formation of microthrombi within the pulmonary vasculature (C) and induces vascular proliferation in the lungs. Consequently, this cascade can culminate in the development of intrapulmonary right-to-left shunt (D). (E) Loss of lung perfusion regulation and hypoxic vasoconstriction might also contribute to right-to-left shunt (D). (F) Respiratory alkalosis and hypocapnia blunt both ventilatory stimuli and sensation of dyspnea.

Abbreviation: P-SILI, patient self-inflicted lung injury.

(0.39MB).

To our knowledge, this study is the largest cohort of silent hypoxemia in COVID-19 to date. Advanced age and the presence of comorbidities emerged as significant independent predictors of adverse outcomes, aligning with prior research. Remarkably, a comprehensive meta-analysis assessing the isolated effect of age on the risk of severe outcomes among COVID-19 patients, the risks of hospitalization, in-hospital mortality, and overall mortality increased by 3.4%, 5.7%, and 7.4% per age year, respectively. These outcomes underscore the continuous age-related increase in susceptibility to adverse events associated with SARS-CoV-2 infection.1,3,29–32 Moreover, the association of comorbid conditions, such as cardiovascular disease, diabetes, and obesity, with severe COVID-19 emphasizes the importance of considering the underlying health status in risk stratification and clinical management strategies for patients with silent hypoxemia.33–35

In our cohort, survivors were generally more symptomatic, experiencing a range of symptoms including anosmia, cough, diarrhea, dysgeusia, fatigue, and sore throat. These symptoms may have prompted patients to seek medical care earlier in the disease course; thereby, increasing the likelihood of receiving timely respiratory support. This prompt intervention could have contributed to diminishing P-SILI, attenuating the severity of respiratory compromise and ultimately improving patient outcomes.

Notably, the development of cough and the need for non-invasive respiratory support emerge as protective factors against mortality. These results underscore the significance of early identification of hypoxemic patients. Timely recognition of symptoms like cough and prompt initiation of appropriate respiratory support measures could significantly improve patient outcomes and reduce mortality rates in COVID-19.

In a meta-analysis investigating the prevalence and outcomes of COVID-19 patients with silent hypoxia, Bepouka et al. demonstrated a mortality rate ranging from 1% to 45.4%, with a pooled mortality of 2%.36 In our cohort, the mortality was notably higher at 28.5%. This disparity may stem from variations in patient demographics, disease severity, healthcare systems, and definitions of silent hypoxemia across studies.

Among non-dyspneic COVID-19 patients, approximately one-third may manifest hypoxemia upon admission, marking them as a subset of severe cases requiring early identification.37,38 Pulse oximetry emerges as a pivotal tool in this context, offering a simple, non-invasive, and widely accessible means of assessing arterial oxygen saturation.39,40 Despite potential confounding factors, such as racial bias, vasoconstriction, hypoxemia, and COVID-19 itself, which may impact the accuracy of SpO2 measurement and the correlation between SpO2 and arterial oxygenation, the integration of pulse oximeters is crucial for screening COVID-19 patients.41,42 In this specific group of non-dyspneic patients, the initial assessment of SpO2 through pulse oximetry holds particular significance, enabling the early detection of asymptomatic hypoxia, facilitating timely intervention, and ultimately enhancing outcomes.37,43,44 Additionally, clinical evaluations should emphasize monitoring the progression of signs and symptoms, especially the respiratory rate, to promptly identify any deterioration.16,45 This comprehensive approach ensures a proactive and effective management strategy for this specific cohort, contributing to enhanced patient care and better overall prognosis.

Regarding screening for the severity of disease in the outpatient setting, it is clear that the sole utilization of patient-reported symptoms would not be an accurate and thorough strategy due to the existence of non-dyspneic hypoxemic presentations. This should warrant the utilization of more objective criteria, including arterial blood gas and pulse oximetry. Based on these findings, we suggest that when feasible patients with COVID-19 should undergo timely screening of hypoxemia, regardless of symptomology, with pulse oximetry to assess for silent hypoxemia and consequent need for respiratory support and pharmacological treatments available.

When providing respiratory support to patients with silent hypoxemia – believed to be associated with the type L phenotype –, the primary focus should be on preventing P-SILI and increasing the fraction of inspired oxygen. Implementing an early strategy of providing respiratory support for these patients could have the potential to reduce the risk of P-SILI, and, consequently, mitigate disease progression.

This study has limitations. First, we used administrative records from structured surveillance forms, which may carry inherent limitations due to their reliance on predefined categories and standardized data entry. Second, not all cases of COVID-19 are notified in Brazil. The underreporting could impact the generalizability of the study results. Third, patients with missing data regarding SpO2 and the presence of dyspnea or respiratory distress at hospital admission were excluded from the analysis. Fourth, since our study was conducted in early 2021, patients with the Omicron variant and its sublineages were unlikely to be included. Thus, further research specifically including patients infected with these variants is essential. Fifth, the absence of available data on laboratory results, including arterial blood gas, which could be used to analyze better blood gas alterations that could potentially interfere with clinical courses such as hypo and hypercapnia, limited our considerations. Sixth, the limited data on time from the symptom onset to the initial presentation and duration of follow-up hinders precise assessment of patient clinical evolution and the impact of delayed medical intervention. Thus, more studies with comprehensive time-related data are warranted to further elucidate the clinical course of silent hypoxemia in COVID-19.

In this multicenter cohort study of 46,102 hospitalized patients due to SARS due to COVID-19 with silent hypoxemia at presentation in Brazil, advanced age, presence of comorbidities, and development of respiratory distress were independent risk factors for mortality. Conversely, the emergence of cough and the requirement for non-invasive respiratory support were independent protective factors against mortality.

This study was sponsored by the Brazilian Council for Scientific and Technological Development (Conselho Nacional de Desenvolvimento Científico e Tecnológico,CNPq, in Portuguese) and by the Coordination of Improvement of Higher Education Personnel (Coordenação de Aperfeiçoamento de Pessoal de Nível Superior, CAPES, in Portuguese). The funding sources did not participate in study design; collection, management, analysis, and interpretation of data; and preparation, review, and approval of this manuscript.

The authors have declared no conflicts of interest.

All data relevant to the study are included in the article or uploaded as supplementary information.