We found a statistically significant moderate to strong negative correlation between the SF ratio and not-, and poorly-inflated lung ratios in critically ill patients with SARS-CoV-2 pneumonia. Actually, an increased not- and poorly-inflated lung, FWHM, lung density and a decreased SF ratio were observed in critically ill patients when compared to inpatients. During the SARS-CoV-2 pandemic, it is important to identify the critically ill patient, thus, CT findings could be used to evaluate the lung heterogeneity and to define disease severity. However, SF ratio is an easier and cheaper noninvasive method when compared to CT, as its overuse during the SARS-CoV-2 pandemic raises concerns about radiation-induced adverse health effects, both in patients and in healthcare workers.
Lung heterogeneity are associated with ARDS severity and independently associated with outcome.11 Different CT scoring systems were presented to estimate the lung heterogeneity and to define disease severity. Our results indicated that increased FWHM in patients with viral pneumonia could be used to estimate the lung heterogeneity. In addition, the current study elicited a significant moderate negative correlation between the SF ratio and FWHM in these patients. Ragap et al. found a strong negative correlation between a high CT severity score by HRCT and oxygen saturation (r = − 0.73, p = .001).12 Similarly, Wang et al. reported a significant moderate negative correlation between oxygen saturation and CT findings (r = − 0.446, p < 0.05), but heterogeneity index was not used.13 Unfortunately, both authors did not report as to when the timing of the values of oxygen saturation were obtained, i.e. on the same day with chest CT or not. In addition, they did not clarify whether the oxygen saturation values were recorded while using oxygen support therapy. 12,13
It has been previously reported that there is a strong and significant association between the square root of the SF ratio value and the risk for death, with a unit decrease in the marker corresponding to a 1.82 fold increase in mortality risk (95% CI: 1.56–2.13). Thus, the SF ratio can be used to define disease severity and mortality risk.14 Optimal cut-off point of SF ratio to identify critically ill patients with viral pneumonia is not found in the literature. In our cohort, a SF ratio cut-off < 445 identified critically ill patients. Interestingly, when the cut-off is taken as < 325, which is close to the ARDS limit-, its selectivity decreases. It is important to evaluate patient breathing effort in addition to SF or PF ratio in the decision of intubation in critically ill patients with viral pneumonia. Bauer et. al. showed that delayed intubation (after two days) was associated with a greater risk of death after 10 days of ICU stay, than early intubation. 15
ICU mortality rates reported in patients with severe SARS-CoV-2 range from 20–62%. In our study, most of the patients with SARS-CoV-2 pneumonia admitted to the ICU with acute respiratory failure survived; the cumulative mortality rate was around 29% in the ICU. Despite the global efforts to reduce it, ICU mortality remains high and varies from country to country, and across different regions within each country, even between hospitals in the same region depending on the level of experience in the managing of critically ill patients. Similar to previous reports, our critically ill patients were typically older, male gender and more sedentary than inpatients.16 It is well known that the majority of older adults are sedentary. A large cohort study showed that consistently meeting PA guidelines was strongly associated with a reduced risk for severe COVID-19 outcomes among infected adults.17 PA is an important and modifiable risk factor for severe COVID-19 outcomes according to the results of this study.
Our results indicated that lower lobes were more affected than the middle and upper lobes. Indeed, inpatients had more well inflated lung ratios. These lung ratios also showed a significant negative correlation with age and BMI. In contrast, critically ill patients had more not- and poorly-inflated lung ratios, and did not correlate with age and BMI. This could be explained by positive pressure ventilation, applied NIV or invasive mechanical ventilation, and limiting or preventing atelectasis especially in older and obese critically ill patients. The application of supraphysiological PEEP (> 5 cmH2O) with invasive mechanical ventilation decreases atelectasis, improves alveolar ventilation and V/Q mismatch.
This study has several limitations. First, only 87 patients with a positive of RT-PCR were included, 169 patients with a negative of RT-PCR were ruled out in the analysis. The study started on March 20, when the first patient was admitted to the COVID-ICU. It was conducted until April 20, 2021 when immunomodulatory treatments began. Patients who did not receive immunomodulatory therapy were included; therefore there is a time limitation (only 1 month). The degree of PA was determined subjectively by referring to statements made by patients or relatives. In addition, optimal inspiration was not achieved in some CT scans especially in the ICU patients. Inadequate inspiration may falsely result in higher parenchymal CT density. Moreover, some CT scans could not be segmented or accurately segmented by the radiological software. Although manual corrections were performed in inadequately segmented scans, this could not be achieved in cases that segmentation failed. Since patients usually had a single CT during their hospitalization, correlation analysis could not be performed between the findings of CT and the worst P/F and SF values in the ICU. The value of the SF ratio as a predictor to determine the need for ICU admission may be high. Patients who might need intubation were admitted to the ICU for close monitoring.