The Effects of Steroids in Severe Hospitalized Patients with COVID-19: A Retrospective Cohort Study


 The efficacy of corticosteroids in the treatment of patients with severe COVID-19 remains unknown. We evaluated the impact of corticosteroids on clinical improvement among severe COVID-19 patients. In this retrospective, two-centered, cohort study, we enrolled 101 patients with severe COVID-19: with 39 patients in the steroid group and 63 patients in the non-steroid group. The primary endpoint was Time to Clinical Improvement (TTCI) by up to 28 days after the treatment. Secondary endpoints included the rate of CAT scan improvement, the percentage of negative SARS-Cov-2 RT-PCR tests by Day 28, and the time to discharge. We found that patients in the steroid group did not have significant differences of TTCI from patients in the non-steroid group by 28 days after the treatment (median, 19 days vs. 20 days; hazard ratio, 1.07; p=0.797). The CAT scan improvement rate was not statistically different between the two groups by Day 28 (87.2% vs. 79.0%, p=0.170). The negative test of SARS-CoV2 RT-PCR by Day 28 was 68.4% in the steroid group, 87.1% in the non-steroid group (p= 0.060). Time to discharge was significantly longer in the steroid group than the non-steroid group (35 days vs 21 days, p=0.005). Our findings indicated the short-term corticosteroid at a low to moderate dose did not improve the clinical outcomes for patients with severe COVID-19. Further randomized clinical trials are needed to confirm the findings.

cases of SARS and in uenza A pneumonia. 18,19 Considering the toxicities involved with high doses of steroids, the expert consensus from the Chinese Thoracic Society stated that a low to moderate dose of steroids (≤0.5-1mg/kg per day methylprednisolone or equivalent) for a short duration (≤7 days) has the probability of reducing mortality in critically ill patients. 15,20 This retrospective study evaluates the e cacy of corticosteroid at a low to moderate dose in the treatment of non-Intensive Care Unit (ICU) patients with severe COVID-19 in Wuhan, China.

Patients and demographics
We reviewed the medical records of 760 hospitalized patients diagnosed with COVID-19 in two study hospitals in Wuhan, of which 659 patients were excluded because they had a mild to moderate disease.
In total, 101 patients were included for data analysis, with 39 patients in the steroid group and 63 in the non-steroid group. Eligible patients were transferred from other hospitals or were admitted directly from the outpatient clinic. The average time from disease onset to the admission was 12 days. The median age was 68, and 81 (77.9%) patients were older than 60 years. 45 (43.6%) patients were men and 3 (3.0%) were smokers. Many patients had comorbidities such as diabetes (30.7%), hypertension (51.5%), cardiovascular disease, including CAD, arrhythmia, and CHF (24.9%). Positive tests for SARS-Cov2 RT-PCR were seen in 73 (70.2%) patients; positive IgM/IgG tests were seen in 66 (65.3%) patients. Some parameters were statistically different between two groups, including symptoms of shortness of breath, SpO2, higher WBC or neutrophils counts, and lower lymphocyte count. (Table 1)

Primary and Secondary Endpoints
For the primary endpoint, we rst performed the univariate analysis using the Kaplan-Meier method. The median TTCI during the 28-day study period was 19 days in the steroid group, similar to 20 days in the non-steroid group (p=0.856). (Table 3, Figure 1). Further multivariate analysis with Cox Professional Hazards model showed no signi cant difference in TTCI between patients in the steroid group and those in the non-steroid group (hazard ratio for clinical improvement, 1.07; 95% con dence interval [CI] 0.62-1.86, p=0.797). (Table 5). When we evaluated the number of patients that achieved clinical improvement with a 2-point decrease in the score on Day 7, Day 14 and Day 21 after the treatment, there was no signi cant difference between the two groups on Day 7 and Day 21(p=1.000 and p=0.243 respectively).
However, more patients in the steroid group achieved a 2-point decrease than those in the non-steroid group at Day 14 (p=0.015), suggesting a transient bene t. (Table 3) For the secondary endpoints, the CAT scan improvement from the baseline was observed in 87.2% of patients in the steroid group by day 28, while CAT scan improvement was observed in 79.0% of patients in the non-steroid group. The difference was not statistically signi cant (p=0.170). The median time to CAT scan improvement from the admission was 16 days in the steroid group, longer than 14 days in the non-steroid group. The percentage of negative SARS-CoV2 RT-PCR results by day 28 was 68.4% in the steroid group, trending signi cantly lower than that (87.1%) in the non-steroid group (p= 0.060). Time to discharge was signi cantly longer in the steroid group than the non-steroid group (36 days vs 21 days, p=0.0005). (Table 3)

Risk Factors
To evaluate whether factors other than steroid treatment could impact clinical improvement, we performed a univariate analysis with a log-rank test on other treatments and lab parameters. Treatment (Arbidol and Expectorants) and abnormal high laboratory ndings (WBC, LDH, IL-6, D-dimer) signi cantly extended the time to clinical improvement. However, the time to clinical improvement was much shorter in patients with high levels of hemoglobin (7.0±5.0 days) than in patients with normal (20.0±1.0 days) or low levels of hemoglobin (20.2±1.2 days) (p=0.004) ( Table 4). All of the factors above except abnormally high WBC remained statistically signi cant as independent factors associated with the time to clinical improvement through multivariate analysis (Table 5).

Adverse effects
In terms of the side effects of corticosteroid, hyperglycemia was detected in a signi cantly higher percentage of patients in the steroid group (21, 53.9%) than in the non-steroid group (15,23.4%) (p=0.003). No patient developed psychosis or avascular necrosis during the hospitalization. (Table 6). Guidance on the use of corticosteroids in the treatment of patients with COVID-19 has been based on limited e cacy and safety data, which is mostly driven from studies in other respiratory viral infections such as SARS, MERS, and in uenza. In a retrospective study, investigating the risk factor of SARS mortality, high-dose corticosteroid was found to increase 30-day mortality. 21 In another small randomized trial including 16 non-critical patients with SARS, hydrocortisone use in 9 patients led to greater viremia in the second and third weeks after infection as compared to the 7 patients receiving placebo. 11 Three other retrospective studies in SARS patients found that corticosteroid use was associated with signi cant toxicities including steroid-induced diabetes, psychosis, and avascular necrosis. 8,10,12 Based on the results above, WHO and Russel C discouraged systemic corticosteroid treatment of COVID-19 patients. 16,17 However, there has been published data that supports the use of corticosteroids in the treatment of viral pneumonia. In a retrospective chart review study, 17 patients receiving high-dose pulse corticosteroids followed by tapering had fewer oxygen requirements and a better imaging outcome than the 55 patients who received low doses of corticosteroids. 22 Another study showed that a low-dose of methylprednisolone (40-80mg daily) was associated with shorter hospital stay than high-dose methylprednisolone (320-640mg daily). 23 In a large retrospective study involving 401 SARS patients, the use of corticosteroids on critical patients resulted in a lowered mortality and a shorter hospital stay in critical patients. The author suggested that the steroid use should not be delayed until Acute Respiratory Distress Syndrome (ARDS) is fully developed. 19 Preliminary data showed that corticosteroids reduced in ammation in 15 critical COVID-19 patients, although there was no mortality improvement. 15 Considering the potential bene ts and risks of corticosteroids, especially when given in high doses (>150mg prednisone or equivalent per day), experts in China formed a consensus to recommend low dose, short-term steroids in selected patients, most notably among patients with severe COVID-19. 7,20 We studied clinical outcomes in a cohort of COVID-19 patients with severe symptoms and systemic in ammation as observed by the elevation of cytokines such as IL-6, CRP, or D-dimer, but without ARDS or secondary hemophagocytic lymphohistiocytosis (sHLH) induced multiorgan failure seen in critical condition. 24 Our data reveals that a short course of corticosteroids, at low to moderate dose did not signi cantly improve clinical symptoms and oxygen requirements by Day 28 of the treatment, although it provided the transient bene t on Day 14. Moreover, the use of corticosteroid prolonged the hospital stay by up to 15 days and failed to improve viremia clearance. These results were consistent with recent small studies, suggesting that the steroid interferes with the ability of the distressed immune system to ght against viruses. 11,25 The lack of bene ts found in corticosteroid treatment is also possibly due to its incomplete inhibition of excessive in ammation in COVID-19 patients. Corticosteroids could not lower interleukin-6 (IL-6) signi cantly in SARS and COVID-19 patients. 15,26 In this study, we found IL-6 is an independent risk factor negatively associated with the clinical improvement, consistent with the nding that IL-6 is a predictor of COVID-19 fatality. 27 It is possible that IL-6 and IL-6 induced cytokine storm syndrome play an important role in the pathophysiology of COVID-19. 24 Indeed, preliminary studies showed that tocilizumab, an IL-6 receptor monoclonal antibody, was potentially effective in the treatment of patients with COVID-19 when used with or without steroids. 28,29 A multicenter, randomized clinical trial of tocilizumab in patients with severe COVID-19 is currently ongoing to address the anti-in ammatory treatment.
Besides IL-6, we found several other independent factors associated with the clinical improvement in patients with severe COVID-19. LDH and D-dimer were signi cantly higher among patients with severe disease than those with milder disease and were associated with poor prognosis. Interestingly, elevated HgB was signi cantly associated with clinical improvement. The underlying immunopathogenic mechanism is unknown. We also observed that treatment with Expectorant and Arbidol, an antiviral drug, was associated with extended time to clinical improvement. However, these observations were not the primary endpoint. Further study is needed to evaluate their potential associations with clinical outcome of COVID-19 infection.
There are several limitations to this study. As a retrospective chart review, we cannot address and balance all potential confounding factors completely. The exact reason for using corticosteroids has not always been recorded in the medical records. More patients in the steroid group were treated with HFNC or NIV than patients in the non-steroid group (8 vs. 1), indicating that some patients were in a more severe condition in the steroid group. The dose of corticosteroids varied among different patients. Finally, patients in this study had severe COVID-19 disease, but they were not critically ill, so the ndings may not be generalized to patients in ICU.

Conclusions
In summary, our study demonstrated that the short-term corticosteroids at low to moderate doses did not improve the clinical outcomes for patients with severe COVID-19. Further randomized clinical trials are needed to con rm this nding.

Study Design and Participant Eligibility
This multiple-center, retrospective cohort study was conducted at two campuses of Wuhan Tongji Hospital, a designated hospital for treating patients with COVID-19. We reviewed charts of patients who were treated from February 11st, 2020 to March 27th, 2020. Patients were eligible for the study if they were at least 14 years old, with con rmed COVID-19 according to WHO and Chinese guidelines, and had a severe case of the disease. 7 Severe patients were de ned as those who had at least at least one of three clinical criteria: PaO2/FiO2 ≤300mmHg, RR≥30/min, SpO2 ≤ 93%, or imaging study showed area of infection enlarged by 50% or more within 24-48 hours. Patients in the steroid group needed to receive the steroid for more than 24 hours to be eligible. Patients were excluded if they were < 14 years old, pregnant or postpartum, or had mild, moderate or critical COVID-19. The research was approved by the ethics commission of Xiang'an Hospital of Xiamen University and the Third Xiangya Hospital of Central South University. Written informed consent was waived due to the characteristics of this retrospective chart review study and the rapid spread of COVID-19.

Data Collection
Data were collected retrospectively from the electronic medical records, imaging studies and laboratory records, medication administration record (MAR) and nursing care documents to ensure accuracy and timely completion. Collected data were reviewed by two independent researchers. Any missing information or answers to questionable data was provided by the providers who were directly involved in patient care.

Outcomes
The primary endpoint was Time to Clinical Improvement (TTCI) by up to 28  All methods were carried out in accordance with relevant guidelines and regulations.

Consent for publication
Not applicable Availability of data and materials The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.

Competing interests
The authors declare that they have no competing interests