New research suggests that detection of pleural effusion on computed tomography (CT) scans significantly elevates intensive care unit (ICU) admission, mechanical ventilation use and 30-day mortality risk in patients with COVID-19.
For the retrospective multicenter study, recently published in Clinical Imaging, researchers reviewed chest CT findings and outcome data from 1,183 patients with COVID-19 (median age of 63) drawn from 13 facilities in Germany. According to the study, 70.7 percent of the cohort were men and 29.3 percent were women.
Chest CT findings revealed pleural effusion in 31.5 percent of patients. The researchers found that patients with pleural effusion were over four times more likely to require mechanical ventilation with a multivariable hazard ratio (HR) of 4.48, and over three times more likely to be admitted to intensive care units (ICUs) with a multivariable HR of 3.10.
When assessing 30-day mortality rates, the study authors noted that pleural effusion had over a 20 percent higher incidence in non-survivors (47.5 percent vs. 27.3 percent). Multivariable analysis revealed that pleural effusion was associated with a 39 percent higher 30-day mortality risk.
“These findings suggest that the detection of (pleural effusion), irrespective of its volume or density, should be a central component in the radiological evaluation of COVID-19 patients to aid in risk stratification and inform clinical decision-making,” wrote study co-author Alexey Surov, M.D., who is affiliated with the Department of Radiology and Nuclear Medicine at Ruhr University of Bochum in Bochum, Germany, and colleagues.
The researchers noted no significant difference in pleural effusion density between non-survivors and survivors (2.6 HU vs. 2.4 HU). However, they did point out that non-survivors had larger total pleural effusion volume than survivors (5,417 mm3 vs. 1,461 mm3).
Three Key Takeaways
- Increased risk with pleural effusion. Patients with pleural effusion seen on chest CT scans were over four times more likely to require mechanical ventilation and over three times more likely to need ICU admission, highlighting the condition as a significant risk factor in COVID-19 outcomes.
- Higher 30-day mortality risk. Pleural effusion was associated with a 39 percent higher risk of 30-day mortality, and its incidence was more common in non-survivors than survivors (47.5 percent vs. 27.3 percent), underscoring its potential role in predicting patient prognosis.
- Prognostic value in COVID-19. The detection of pleural effusion should be integrated into the radiological evaluation of COVID-19 patients as it has been shown to be a strong prognostic indicator, independent of traditional lung consolidation markers.
Noting that pleural effusion was initially considered an uncommon finding in patients with COVID-19, the study authors emphasized that the presence of this condition in critically ill patients and multivariable analysis findings from this study suggest that pleural effusion is a significant consideration.
“The independence of PE as a prognostic factor, independent of traditional markers of lung consolidation, underscores its importance in the radiologic evaluation of COVID-19 and highlights its relevance for consideration in the diagnostic and management pathways of this disease,” maintained Surov and colleagues.
(Editor’s note: For related content, see “COVID-19 and Cancer: What a New Chest CT Study Reveals,” “New Chest Imaging Study Finds Intrathoracic Complications in 20 Percent of Patients with COVID-19” and “What a New Chest CT Study Reveals About Pericardial Effusion and COVID-19.”)
In regard to study limitations, the authors noted the retrospective study design and the lack of an international study arm. While it was a multicenter study, the researchers acknowledged the data was derived from patients treated at 12 tertiary care centers during the COVID-19 pandemic. Accordingly, the authors cautioned against extrapolation of the study findings to broader patient populations from primary care, secondary care, and non-tertiary facilities. The readers also suggested the possibility of bias due to the lack of central reading in this study for cross-validation.
