In this retrospective monocentric study, which focused on the second and third waves of the pandemic in Germany, slightly more than two thirds of the hospitalized patients with COVID-19 (68%) received specific drug therapy, with the proportion of patients without specific therapy halving from approximately 40% to 20% from the second to the third waves. Among the drug therapy approaches, corticosteroids (primarily dexamethasone) were used most frequently, accounting for more than 60% of the cases. Among these, the proportion of steroid administration increased from 56 to 75% of the patients from the second to the third waves, with a concomitant decrease in steroid monotherapy to one third of the patients and a doubling of the proportion of combination therapy with remdesivir during the third COVID-19 wave to 40%.
The results of clinical trials, systematic reviews and meta-analyses that investigated the administration of remdesivir in hospitalized patients remain inconsistent and continue to provide uncertainty regarding its potential benefits [20,21,22,23,24,25,26,27,28,29]. A recent meta-analysis of the efficacy and safety of remdesivir  summarized the clinical recovery and all-cause mortality end points from three RCTs [4, 30, 31] and one observational study . According to this analysis, the rate of recovery with remdesivir therapy increased by 21% (risk ratio [RR] 1.21 [95% CI 1.08–1.35]) at day 7, by 29% (RR 1.29 [95% CI 1.22–1.37]) at day 14, and by 9% (RR 1.09 [95% CI 1.01–1.14]) at day 28; pooled estimates of all-cause mortality showed a 39% reduction in mortality risk with remdesivir therapy at day 14 (RR 0.61 [95% CI 0.46–0.79], P = 0.003) and a non-significant 22% reduction in all-cause mortality at day 28 (RR 0.78 [95% CI 0.59–1.03], P = 0.09). After exclusion of the study by Wang et al.  from the meta-analysis, the pooled data showed a significant 27% reduction in 28-day mortality (RR 0.73 [95% CI 0.54–0.99], P = 0.04) . A recent RCT (PINETREE) of non-hospitalized patients at high risk for COVID-19 progression showed that a 3-day course of remdesivir resulted in an 87% lower risk of hospitalization or death than placebo . In contrast to the PINETREE trial, which involved early administration of remdesivir in the outpatient setting, the randomized controlled DisCoVeRy trial found no clinical benefit from the use of remdesivir in patients who were hospitalized for COVID-19, had symptoms for more than 7 days and required oxygen support . By contrast, there is reliable evidence of the clinical efficacy of dexamethasone in hospitalized patients with moderate to severe COVID-19. Compared to the group of patients receiving SOC only, a significant reduction in 28-day mortality was observed in patients with dexamethasone therapy and invasive ventilation (29.3% vs. 41.4%, RR 0.64 [95% CI 0.51–0.81]), and in those with non-invasive ventilation therapy (23.3% vs. 26.2%; RR 0.82 [95% CI 0.72–0.94]) . However, no data from RCTs are available yet for the combination of corticosteroids with remdesivir [13, 34,35,36,37].
This prompted us to conduct a comprehensive retrospective analysis of the efficacy of combination therapy with corticosteroids and remdesivir compared with remdesivir monotherapy in our own setting. Clinical improvement occurred within 28 days in 69.4% of the patients receiving corticosteroids and in 75.5% of those in the corticosteroid-remdesivir group. The median time to clinical improvement of 13 days was the same in both matched therapy groups; this was also demonstrated in the subgroup analysis by disease severity (i.e., treatment in normal wards vs. ICU; and low- vs. high-flow oxygen therapy). Mortality was significantly lower in the corticosteroid-remdesivir group at 14.8% compared with 22.2% in the corticosteroid group (P = 0.048). The survival benefit was also clearly evident in the Cox regression analysis (HR 0.60, P = 0.03) for combined corticosteroid-remdesivir treatment. Further differences were seen in the subgroup analysis by disease severity (normal ward: HR 0.45, P = 0.015; low-flow oxygen therapy: HR 0.26, P = 0.007). As disease severity increased (treatment in the ICU or high-flow oxygen therapy), no significant difference in mortality between the two treatment groups could be detected. These results are consistent with those of a prospective, non-randomized clinical trial from Italy of 151 COVID-19 patients (76 cases with remdesivir plus dexamethasone vs. 75 cases with dexamethasone as monotherapy), which showed that combined treatment with remdesivir plus dexamethasone reduced 30-day mortality, shortened the length of hospital stay, and prevented disease progression . A Danish multi-center cohort study showed that the combination of remdesivir plus dexamethasone reduced 30-day mortality compared with standard non-pharmacologic treatment (SOC); the 30-day mortality rate of 1,694 patients receiving additional remdesivir plus dexamethasone decreased to 12.6% compared with 19.7% in 1,053 patients in the SOC group (OR 0.47) .
We also performed a subgroup analysis to evaluate the efficacy of remdesivir alone, without additional administration of glucocorticoids. Clinical improvement was achieved in 88.6% of our patients who received remdesivir monotherapy compared with 68.2% of the patients with SOC only, demonstrating the treatment benefits of remdesivir therapy (OR 3.64, P = 0.024). However, the median time to clinical improvement was comparable at 10 days in both the group without remdesivir treatment and the remdesivir group. The reduction in mortality from 27.3% in the patients without specific therapy to 6.8% in those treated with remdesivir was significant (OR 0.195, P = 0.017). In Cox regression analysis, remdesivir treatment was associated with significant benefits (HR 0.20, P = 0.006). Significant differences in mortality were also seen in the patients remaining in the normal ward (HR 0.22, P = 0.01) and in those receiving low-flow oxygen therapy (HR 0.23, P = 0.04). The mortality rate of the patients receiving corticosteroid plus remdesivir combination therapy was comparable at 15.5% during the third wave and 15.4% during the second wave. By contrast, the mortality rate for corticosteroid monotherapy decreased from 27.9% during the second COVID-19 wave to 12.8% during the third COVID-19 wave. In particular, the administration of methylprednisolone as pulse therapy (IV, 250 mg/day for 3 days) has been established in the treatment of severely ill COVID-19 patients during the third wave of COVID-19 and has led to a significant reduction in mortality compared with standard treatment .
The third COVID-19 wave in Germany was mainly driven by the emerging SARS-CoV-2 alpha variant (B.1.1.7), which has a higher reproductive number compared with the wild type virus . For this reason, the federal government in Germany extended nationwide lockdown measures until early May 2021, resulting in a uniform decline in the number of infections in all states . These measures and the vaccination of the most vulnerable groups, particularly the elderly, against SARS-CoV-2, which began in late December 2020, resulted in the containment of rising infection rates during the third wave of COVID and, consequently, a decrease in the hospitalization rates of the most vulnerable groups. Another striking feature of hospitalized patients during the third COVID-19 wave was the marked decrease in chronic diseases, such as hypertension, cardiovascular disease, diabetes, renal disease, and dementia, being associated with potential complications of COVID-19 and increased disease severity [41,42,43,44]. The likelihood of hospitalization because of COVID-19 was very high in patients with pre-existing conditions. The known risk factors for severe disease are older age, hypertension, cardiovascular and chronic lung disease, diabetes, immunodeficiency, smoking, and male sex [42,43,44,45,46]. During the study period, 74.1% of the patients with COVID-19 were admitted with cardiovascular disease. Among them, the diagnosis of hypertension was very frequent (67.3%). Metabolic disorders, such as diabetes (30.5%), obesity (15.4%), and renal disease (30.9%) also dominated. Although 65.1% of the patients had mild or moderate COVID-19 (WHO scores 3–4), there was a high proportion of patients with severe (scores 5–7) and fatal courses (score 8). Mortality in our study population was 18.2%, which is comparable to in-hospital mortality in other tertiary care centers in France (18.1%)  and Germany (17%) , but significantly lower than that in the UK, where an in-hospital mortality of 26% has been reported .
Our study has several limitations. First, it is a retrospective, observational study at a single center, and despite the use of propensity score matching, we cannot exclude the possibility that there are other unmeasured confounders that could bias our estimates of the treatment effect. For example, remdesivir has to be administered intravenously, for which there are hardly any outpatient structures in Germany, and some of the patients did not receive remdesivir because the disease was already too advanced. Late presentation may have influenced selection towards a severe course in several ways. Second, despite the large size of the cohort studied (n = 839 fully evaluable patients), subgroups are relatively small, especially when the analysis was restricted to matched groups, and the validity of our results needs to be confirmed in larger studies. However, although the results of a retrospective cohort cannot replace the results of RCTs, these data may help answer a question for which RCT data are incomplete, namely whether the use of remdesivir is associated with lower mortality in hospitalized patients with COVID-19.