Demographics and baseline characteristics of patients with severe COVID-19
98 hospitalized patients, laboratory confirmed to be SARS-COV-2 positive, were included in the study. The average age for the participants was 45.8 years (Std.Error (SE), 1.4). Of these, 58 (59.2%) were male. 76 (78.6%) of the participants were considered to be moderate cases, while 22 (22.4%) were considered to be severe cases (Table 1). The median duration from symptom onset to hospital admission, hospital length of stay (HLOS) and clinic course were 6 days (IQR, 5), 15.6 days (IQR, 11), and 21.7 days (IQR, 9), respectively. Of the 98 patients, 32 (32.7%) had 1 or more comorbidities. Hypertension [13 (13.3)], diabetes [6(6.1)], cancer [4(4.1)], and chronic liver disease [3(3.1)] were the most common underlying conditions (Table 1).
Compared with moderate cases, the severe patients were significantly older [average age, 55.1 years (SE, 2.2) vs 43.1 years (SE, 1.5), P < 0.001]. Relative to moderate cases, severe cases were associated with longer HLOS and clinic course [median, 20.5 days (IQR, 8.5) vs 13 days (IQR, 11.0), P = 0.001; 26.5 days (IQR, 14.3) vs 19.5 days (IQR, 8.8), P<0.001, respectively]. Relative to moderate cases, significantly more severe patients were treated with antibacterial agents, glucocorticoids and immunoglobulin [12(54.5%) vs 20(26.3%), P=0.013; 17(77.3%) vs 4(5.3%), P<0.001; 14(63.6%) vs 2(2.6%), P<0.001, respectively]. There were no differences in the percentage of vitamin C therapy given to moderate and severe cases [7(31.2%) vs 17(22.4%)] (Table 1).
Deterioration of lymphocytopenia and neutrophilia at the lowest oxygenation index timepoint
COVID-19 symptoms continued to progress after hospital admission. During hospital stay, the patients underwent multiple oxygenation index assays, a key lung function indicator. The lowest oxygenation index was chosen as the most severe timepoint. The laboratory results for tests done at admission, at the most severe timepoint and before discharge were analyzed. This analysis revealed that laboratory results for moderate cases remain relatively stable between the admission and the lowest oxygenation timepoint. At admission, the lymphocyte ratio in severe cases was significantly lower than in moderate cases [median, 23.6 (IQR, 20.85) vs 25.75 (IQR, 9.8), P =0.041], while the neutrophil ratio was higher in severe cases [median, 66.1 (IQR, 27.25) vs 65 (IQR, 14.43), P =0.039]. At the most severe timepoint, the lymphocyte ratio in severe cases decreased sharply, and was lower than the lymphocyte ratio in moderate cases [median, 11.55 (IQR, 15.58) vs 25.5(IQR, 13.95), P <0.001]. Neutrophil ratio and white-cell counts were significantly higher than in moderate cases [median, 81.25 (IQR, 21.8) vs 66.1(IQR, 16.12), P <0.001; 8.05 (IQR, 4.93) vs 6.0(IQR, 2.05), P=0.006, respectively] (see Table 2).
At hospital discharge, oxygenation index, lymphocytopenia and neutrophilia were significantly improved after multiple rounds of treatment (P<0.001), when compared to the most severe timepoint. However, neutrophil ratio and white-cell counts in severe cases remained higher than in moderate cases [median, 66.3 (IQR, 15.38) vs 62.85 (IQR, 11.9), P=0.002; 6.0(IQR, 1.7) vs 6.95(IQR, 3.83), P<0.001, respectively]. Notably, the platelet count in severe cases was significantly lower than in moderate cases [median, 205.5 (IQR, 100.25) vs 270 (IQR, 100.5), P=0.003] (see Table 2).
Liver and kidney functions were monitored by analyzing alanine and aspartate aminotransferase levels, as well as creatinine clearance rate. This analysis revealed that liver and kidney function in the severe cases were poorest at admission but improved gradually. Relative to the moderate cases, lower levels of uric acid, an endogenous free radical scavenger and a major antioxidant in plasma[3], were observed in severe cases at the lowest oxygenation index timepoint [median, 224 (IQR, 128) vs 257 (IQR, 139.5), P=0.028] (see Table 2). However, linear regression analysis suggests that the lower uric acid levels might be age and sex related and not oxygenation index dependent (uric acid = 468.65 -1.721age -56.244sex -0.07 oxygenation index, Adjusted R Square =0.098, P=0.036, 0.009, 0.295 for age, sex, oxygenation index, respectively).
Association of oxygenation index with lymphocytopenia and neutrophilia
To evaluate the relationship of oxygenation index with lymphocytopenia and neutrophilia, the pearson correlation analysis and linear regression analysis were carried out. The association of oxygenation index with lymphocytopenia and neutrophilia were demonstrated by pearson correlation analysis of the values at the most severe timepoint (r= 0.465 with neutrophil ratio, P<0.001; r=-0.461 with lymphocyte ratio, P<0.001) (Fig 1).Linear regression analysis showed that age and liver function were the factors that influenced the oxygenation index levels (oxygenation index=790.777-39.816age-2.799sex-1.627body weight-2.385ALT+0.536 AST-0.404 creatinine clearance rate-2.672 comorbidities, Adjusted R Square =0.141, P=0.006, 0.043, 0.226, 0.299, 0.739, 0.563, 0.912 for age, ALT, sex, body weight, AST, creatinine clearance rate, comorbidities, respectively). Syphilis, gout, hypothyroidism were not included in comorbidities here as neither of them was reported to be related with the risk of COVID-19 in the literature.