Baseline Characteristics of Participants in the Study
The clinical characteristics of all patients were summarized in supplement Table 1. Among 452 COVID-19 patients, the average age was 57years old. The number of male (n=225, 49.78%) was similar with female (n=227, 50.22%). The median days lag from illness onset to hospital admission were 10 days (interquartile range, 7-15 days). The nucleic acid negative conversion time was 28 days (interquartile range 21-34 days). Common underlying concomitant diseases included hypertension (16.15%) and diabetes (8.85%).
On admission, all patients had had severe symptoms and received standard care, including supportive and antiviral treatment. Arbidol was the most frequently used antiviral regiments (94.91%). 71.90% of patients received LHQW treatment (four capsules each time, three times daily for two weeks). Dexamethasone (10mg/day, 3-5 days) was used shortly to inhibit inflammatory cascade in 23.23% of patients who expressed progressing disease. Most patients (88.94%) were treated with a single antibiotic, which generally covered common pathogens and some atypical pathogens; when secondary bacterial infection occurred, medication was ministered according to bacterial culture and drug sensitivity.
Since all patients with symptoms of respiratory distress, they received oxygen therapy, including nasal catheters (98.01%), mask (0.89%), and high-flow nasal cannula oxygen therapy (0.22%), or non-invasive ventilator (0.88%) on admission. 13.50% of participants were transformed from nasal catheters to mask, and 1.33% of the subjects were transferred to non- invasive-ventilator with illness worsens. As of March 13, 2020, all patients were survived and discharged.
The Different Characteristics Between Patients with Dexamethasone and without Dexamethasone
Of the 452 patients, 105 (23.23%) subjects received dexamethasone, and 347 (76.77%) did not. The distribution of the patients’ baseline characteristics according to dexamethasone exposure was shown in Table 1, both in the unmatched and propensity-score–matched analytic samples.
There were no differences in age and sex in the unmatched patients with and those without dexamethasone. In contrast to the counterpart, however, dexamethasone-treated patients exhibited a higher percentage of other endocrine system diseases history and higher leukocytes, neutrophile granulocytes, procalcitonin, serum IL-6 levels, D-Dimer and hsTNI, but lower lymphocytes, hemoglobin, albumin at baseline(p<0.05). In CT scan examination, more percentage of dexamethasone-treated patients had lung consolidation shadow. All suggested that patients with dexamethasone exhibited more severe inflection, stronger inflammatory responses, and poorer nutritional status than no dexamethasone-treated- patients. The characteristics led to more percentage of dexamethasone-treated patients receiving anti-virus and supportive management, including Remdesivir, immune globulin, thymosin, plasbumin, mask, and high flow as well as ventilator treatment, when compared those treatments to patients without dexamethasone treatment (p<0.05). Although many differences were aforementioned, the viral RNA shedding had no significant differences between the two groups (p=0.288).
In the matched patients, thirty-three participants were exposed to dexamethasone, and twenty-six patients were not exposed. The differences of variables between patients with dexamethasone and without dexamethasone were attenuated in the propensity-score–matched subjects compared with that in the unmatched samples.
The Correlation Between Dexamethasone Combined with or without LHQW and Nucleic Acid Negative Conversion Time
Table 2 display the association between dexamethasone and nucleic acid negative conversion time. LHQW as an interaction factor affected nucleic acid negative conversion time of patients who were exposed to dexamethasone or were not (p=0.0273). Dexamethasone significantly accelerated nucleic acid negative conversion time than no-dexamethasone (β, -4.77; 95% CI, -9.41, -0.12) in LHQW treated group after adjusted covariable. Corticosteroid using alone had a tendency to prolong nucleic acid negative conversion time than no-corticosteroid using patients (β, 5.37; 95% CI, -4.88, 15.62) in no LHQW treated group, but the result had no significant differences (p=0.3181). Additional multivariable propensity-score analyses yielded consistent results with the unmatched participants (Figure 1)