Corticosteroids with or without antiviral therapy in elderly patients hospitalized with COVID-19 during the Omicron era: A multicenter real-world study

Abstract

Background

Elderly patients remain at high risk for adverse outcomes from COVID-19 despite the generally reduced pathogenicity of the Omicron variant. Corticosteroids are commonly used in hospitalized patients with COVID-19, particularly in those with more severe disease, while antiviral agents inhibit viral replication. However, evidence comparing corticosteroid monotherapy with corticosteroid–antiviral combination therapy in elderly hospitalized patients remains limited.

Methods

This multicenter retrospective cohort study included hospitalized patients aged ≥60 years with laboratory-confirmed COVID-19 at two tertiary hospitals in China, all of whom received systemic corticosteroid therapy during hospitalization. Patients were classified into either a corticosteroid monotherapy group or corticosteroid–antiviral combination therapy group. To reduce confounding, inverse probability of treatment weighting (IPTW) based on propensity scores was applied. The primary outcome was all-cause in-hospital mortality, and mechanical ventilation was assessed as a secondary outcome.

Results

A total of 624 elderly hospitalized patients with COVID-19 who received systemic corticosteroid therapy were included in the IPTW analysis (290 in the corticosteroid group and 334 in the combination therapy group). After weighting, baseline characteristics were well balanced between groups. The weighted incidence of in-hospital mortality was lower in the combination therapy group than in the corticosteroid monotherapy group (8.09% vs 13.47%; OR 0.566, 95% CI 0.333–0.961; p = 0.035). No statistically significant difference was observed in the risk of mechanical ventilation between groups (OR 1.286, 95% CI 0.924–1.789; p = 0.136).

Conclusions

Among elderly hospitalized patients with COVID-19 during the Omicron period who received systemic corticosteroid therapy corticosteroids combined with antiviral therapy were associated with lower in-hospital mortality compared with corticosteroid monotherapy, while no significant difference was observed in mechanical ventilation.

1. Introduction

Coronavirus disease 2019 (COVID-19) can progress to severe or critical illness requiring oxygen therapy or mechanical ventilation and is associated with substantially increased mortality.¹ The World Health Organization (WHO) and several national clinical guidelines recommend systemic corticosteroids for patients with severe or critical COVID-19 to reduce short-term mortality and improve clinical outcomes.^2,3 The RECOVERY trial demonstrated that, among hospitalized patients with COVID-19, dexamethasone reduced 28-day mortality by approximately 36% in those receiving invasive mechanical ventilation, with additional benefit observed in patients requiring oxygen therapy.⁴ Subsequent systematic reviews and meta-analyses have further confirmed that corticosteroid therapy improves clinical outcomes in patients with severe or critical COVID-19 as well as in those with severe community-acquired pneumonia.^2,5,6 In patients requiring mechanical ventilation, corticosteroid therapy has been associated with lower mortality and improved ventilation-related outcomes.^4,7

Antiviral agents have also demonstrated clinical benefits in COVID-19 by inhibiting viral replication. During the Omicron period, small-molecule antivirals such as nirmatrelvir/ritonavir (Paxlovid) became an important therapeutic option, particularly for patients at high risk of disease progression. In a randomized trial, early nirmatrelvir/ritonavir significantly reduced the risk of progression to severe COVID-19 or death in high-risk nonhospitalized adults, and real-world studies during the Omicron wave similarly reported lower risks of hospitalization and mortality with its use8, 9, 10. Because corticosteroids primarily target the dysregulated host inflammatory response, whereas antiviral agents suppress viral replication, combining these therapies may provide complementary therapeutic effects. However, evidence for antiviral therapy in hospitalized patients with COVID-19 remains limited, and data on combination strategies with corticosteroids are still largely derived from observational studies.^1,9,11

Elderly patients hospitalized with COVID-19, particularly those with multiple comorbidities or critical illness, represent a particularly vulnerable population among hospitalized patients with COVID-19, with substantially higher rates of mechanical ventilation and mortality.^12,13 In addition, the long-term outcomes of survivors have received increasing attention. Follow-up studies have shown that patients discharged after hospitalization for COVID-19 may continue to experience persistent symptoms, functional limitations, and reduced quality of life for up to two years after discharge, particularly among those with severe disease or a history of mechanical ventilation.^14,15 Therefore, clarifying the effects of corticosteroid–antiviral combination therapy on hard clinical endpoints such as mortality and mechanical ventilation in patients with severe or critical illness is important for guiding clinical decision-making and informing long-term management strategies.

Currently, evidence from multicenter studies in China that systematically evaluate corticosteroid plus antiviral therapy versus corticosteroid monotherapy in elderly hospitalized patients with COVID-19 remains limited. Moreover, few studies have applied IPTW to adequately balance baseline confounders when assessing these treatment strategies. In this study, we used observational data from two medical centers and applied inverse probability of treatment weighting to balance age, sex, study center, comorbidities, and baseline disease severity. We then compared the effects of corticosteroid plus antiviral therapy versus corticosteroid monotherapy on all-cause mortality and the need for mechanical ventilation, aiming to provide evidence to inform treatment decisions for patients with COVID-19.

2. Methods

2.1. Study design and data source

This retrospective multicenter observational study used real-world data from two tertiary hospitals in China: Nanfang Hospital, Southern Medical University and Huashan Hospital, Fudan University. Consecutive hospitalized patients aged ≥60 years with laboratory-confirmed COVID-19 were screened during the Omicron-dominant period.

Patients from Nanfang Hospital were enrolled between December 2022 and October 2024, while patients from the Baoshan campus of Huashan Hospital were admitted between March 29 and May 17, 2022. The different enrollment periods reflected regional differences in the timing of Omicron-dominant outbreak waves and corresponding hospital admission periods in Guangzhou and Shanghai, rather than an inconsistency in study design or eligibility criteria.

Demographic characteristics, comorbidities, laboratory findings at admission, treatment strategies, and clinical outcomes were extracted from electronic medical records using a standardized form.

2.2. Inclusion and exclusion criteria

Consecutive hospitalized patients aged ≥60 years with laboratory-confirmed COVID-19 who received systemic corticosteroid therapy during hospitalization at the two participating centers were eligible for inclusion. Patients were excluded if they died within 24 h after admission, discontinued treatment during hospitalization, withdrew from active treatment, or were discharged against medical advice.

2.3. Ethics statement

The study was approved by the institutional review boards of Huashan Hospital, Fudan University (protocol number: KY2022-582) and Nanfang Hospital, Southern Medical University (protocol number: NFEC-2023-557). The study was conducted in accordance with the Declaration of Helsinki. Due to the retrospective design and the use of anonymized data, the requirement for informed consent was waived.

2.4. Study population

Eligible patients were categorized into two groups according to their treatment strategy during hospitalization: corticosteroid monotherapy and corticosteroids combined with antiviral therapy. Combination therapy was defined as the use of systemic corticosteroids together with antiviral therapy during hospitalization. The antiviral regimen predominantly consisted of nirmatrelvir/ritonavir (Paxlovid), which was generally administered orally at a dose of 300 mg nirmatrelvir (two 150 mg tablets) plus 100 mg ritonavir (one 100 mg tablet) every 12 h for 5 days. A small number of patients also received other antiviral agents, including simnotrelvir/ritonavir and Azvudine. Because these subgroups were very small, they were not analyzed separately and were included within the overall antiviral therapy category. Systemic corticosteroid therapy included injectable dexamethasone and methylprednisolone administered during hospitalization. The specific agent, dose, and duration were determined by the treating physician according to disease severity and clinical condition. Patients who received corticosteroids without concurrent antiviral therapy were classified as the corticosteroid monotherapy group.

2.5. Outcomes and variables

The primary outcome of interest was all-cause in-hospital mortality. The secondary outcome was mechanical ventilation during hospitalization, reflecting progression to severe respiratory failure. Baseline variables included demographic characteristics (age and sex), hospital center, comorbidities (hypertension, diabetes mellitus, coronary heart disease, chronic obstructive pulmonary disease, chronic kidney disease, and malignancy), and disease severity (severe or critical illness). Baseline laboratory parameters collected at hospital admission included white blood cell count, lymphocyte count, C-reactive protein, procalcitonin, D-dimer, serum creatinine, and estimated glomerular filtration rate. These variables were selected based on clinical relevance and prior evidence suggesting their association with COVID-19 severity and mortality.

2.6. Statistical analysis

Continuous variables are presented as mean ± standard deviation (SD) or median with interquartile range (IQR), as appropriate, and categorical variables are expressed as counts and percentages. Comparisons between groups were performed using the Student's t-test or Wilcoxon rank-sum test for continuous variables and the chi-square test or Fisher's exact test for categorical variables, as appropriate.

To account for potential confounding due to non-random treatment assignment, propensity scores were estimated using a multivariable logistic regression model including age, sex, study center, comorbidities (hypertension, diabetes mellitus, coronary heart disease, chronic obstructive pulmonary disease, chronic kidney disease, and malignancy), and disease severity. Inverse probability of treatment weighting (IPTW) based on the propensity score was then applied to generate a weighted pseudo-population. Stabilized weights were used, and extreme weights were truncated at the 1st and 99th percentiles to reduce the influence of outliers.

Covariate balance before and after weighting was assessed using standardized mean differences (SMD), with an SMD <0.10 indicating adequate balance. Covariate balance was further visualized using a Love plot (Supplementary Fig. S1). After IPTW adjustment, weighted logistic regression models were used to estimate the association between treatment strategy and clinical outcomes. Results were reported as odds ratios (ORs) with 95% confidence intervals (CIs).

Prespecified subgroup analyses were performed as exploratory analyses to assess the consistency of treatment effects across clinically relevant subgroups. Treatment-by-subgroup interactions were evaluated using formal interaction tests, and these analyses were considered hypothesis-generating. All analyses were conducted in the complete-case population with available data for propensity score covariates. Statistical analyses were performed using R software (version 4.5.1; R Foundation for Statistical Computing, Vienna, Austria). A two-sided P value < 0.05 was considered statistically significant.

3. Results

3.1. Patient characteristics

The flowchart of patient selection is presented in Fig. 1. A total of 624 elderly hospitalized patients with COVID-19 who received corticosteroid therapy were included in the primary analysis, including 290 patients receiving corticosteroid monotherapy and 334 receiving corticosteroids combined with antiviral therapy.

Fig. 1.

Flowchart of patient selection. Hospitalized patients with laboratory-confirmed COVID-19 were screened at two participating centers during the study period. Patients were eligible if they were aged ≥60 years and received systemic corticosteroid therapy during hospitalization. Patients meeting the prespecified exclusion criteria were excluded before cohort assembly. A total of 624 patients were included in the primary analysis, including 290 patients in the corticosteroid monotherapy group and 334 patients in the corticosteroids plus antiviral therapy group.

Baseline disease severity varied across the cohort. Before weighting, several baseline characteristics differed between treatment groups. In the unweighted cohort, malignancy was more common in the combination therapy group than in the corticosteroid monotherapy group (24.6% vs. 16.6%), and the distribution of disease severity categories differed between groups.

After applying stabilized inverse probability of treatment weighting with truncation of extreme weights, baseline characteristics were well balanced between the two groups. All covariates included in the propensity score model achieved standardized mean differences (SMD) below 0.10, indicating adequate balance between treatment groups. As shown in Supplementary Fig. S1, absolute SMDs were substantially reduced after weighting and all post-weighting values were below the conventional 0.10 threshold. The weighted distributions of disease severity were similar between groups (non-severe: 33.1% vs. 32.9%; severe: 14.6% vs. 14.9%; critical: 52.3% vs. 52.2%) (Table 1).

Table 1.

Baseline characteristics of elderlyhospitalized patients with COVID-19 before and after inverse probability of treatment weighting (IPTW).

Characteristic	Unweighted Cohort			Weighted Cohort (after IPTW)
	Corticosteroids (n = 290)	Corticosteroids + antiviral (n = 334)	SMD	Corticosteroids (weighted)	Corticosteroids + antiviral (weighted)	SMD
Age, y	73.80 (8.65)	74.86 (9.58)	0.116	74.36 (8.80)	74.37 (9.60)	<0.001
Sex, male, %	192 (66.2)	214 (64.1)	0.045	188 (65.3)	217 (65.1)	0.004
Original comorbidities
Hypertension, n (%)	184 (63.4)	189 (56.6)	0.140	171 (59.4)	197 (59.2)	0.004
DM, n (%)	123 (42.4)	133 (39.8)	0.053	119 (41.2)	137 (41.0)	0.005
CHD, n (%)	80 (27.6)	84 (25.1)	0.055	77 (26.7)	88 (26.3)	0.008
COPD, n (%)	40 (13.8)	44 (13.2)	0.018	38 (13.3)	45 (13.4)	0.002
CKD, n (%)	93 (32.1)	94 (28.1)	0.086	88 (30.3)	100 (30.1)	0.004
Cancer, n (%)	48 (16.6)	82 (24.6)	0.199	59 (20.5)	70 (20.9)	0.009
Disease severity
Non-severe, n (%)	112 (38.6)	95 (28.4)	−0.216	96 (33.1)	110 (32.9)	−0.005
Severe disease, n (%)	36 (12.4)	57 (17.1)	0.132	42 (14.6)	50 (14.9)	0.008
Critical disease, n (%)	142 (49.0)	182 (54.5)	0.111	151 (52.3)	174 (52.2)	0.002

Values are presented as mean (SD) or n (%), unless otherwise indicated. Weighted estimates were generated using stabilized inverse probability of treatment weighting (IPTW) with truncation at the 1st and 99th percentiles. An absolute standardized mean difference (SMD) < 0.10 was considered to indicate adequate balance. DM, diabetes mellitus; CHD, coronary heart disease; COPD, chronic obstructive pulmonary disease; CKD, chronic kidney disease.

3.2. Primary and secondary outcomes

After IPTW adjustment, corticosteroid plus antiviral therapy was associated with a significantly lower risk of all-cause in-hospital mortality compared with corticosteroid monotherapy. The weighted mortality rate was 8.09% in the combination therapy group and 13.47% in the corticosteroid monotherapy group, corresponding to an adjusted odds ratio (OR) of 0.566 (95% CI 0.333–0.961; p = 0.035) (Table 2).

Table 2.

IPTW-adjusted outcomes in elderly hospitalized patients with COVID-19.

Outcome	Corticosteroids (weighted)	Corticosteroids + antiviral (weighted)	OR (95% CI)	P value
All-cause mortality	13.47%	8.09%	0.566 (0.333–0.961)	0.035∗
Mechanical ventilation	36.62%	42.63%	1.286 (0.924–1.789)	0.136

IPTW: stabilized weights, truncated at 1st and 99th percentiles; same cohort and covariates as Table 1. ∗ A two-sided P value < 0.05 was considered statistically significant.

In contrast, no statistically significant difference was observed in the risk of mechanical ventilation between the two groups. The weighted proportion of mechanical ventilation was 42.63% in the combination therapy group and 36.62% in the corticosteroid monotherapy group, with an adjusted OR of 1.286 (95% CI 0.924–1.789; p = 0.136) (Table 2).

3.3. Risk factors for mortality

In the IPTW-weighted regression analysis, combination therapy remained independently associated with lower mortality compared with corticosteroid monotherapy (OR 0.56, 95% CI 0.34–0.94; p = 0.029). Increasing age was also significantly associated with a higher risk of mortality (OR 1.04 per year, 95% CI 1.00–1.08; p = 0.041). Sex, hypertension, diabetes mellitus, and malignancy were not significantly associated with mortality in the adjusted model (Table 3).

Table 3.

IPTW-weighted multivariable analysis of factors associated with all-cause in-hospital mortality.

Variable	OR (95% CI)	P value
Combination therapy(vs corticosteroid monotherapy)	0.56 (0.34–0.94)	0.029∗
Age (per year)	1.04 (1.00–1.08)	0.041∗
Sex (male)	1.48 (0.68–3.25)	0.33
Hypertension	0.62 (0.29–1.32)	0.22
DM	1.59 (0.75–3.39)	0.23
Malignancy	1.02 (0.50–2.10)	0.95

OR, odds ratio; CI, confidence interval; DM, diabetes mellitus. ∗ A two-sided P value < 0.05 was considered statistically significant.

3.4. Subgroup analyses

Exploratory subgroup analyses were conducted according to baseline disease severity, age group, and comorbidity burden. Although effect estimates varied across subgroups, no statistically significant treatment-by-subgroup interaction was observed (Fig. 2). Similarly, exploratory subgroup analyses for the composite outcome (all-cause death or mechanical ventilation) showed no significant heterogeneity in treatment effects across subgroups (Fig. 3).

Fig. 2.

Exploratory subgroup analyses of the association between corticosteroids plus antiviral therapy and all-cause mortality in the IPTW-weighted cohort. Forest plot showing odds ratios (ORs) and 95% confidence intervals (CIs) for all-cause mortality by prespecified subgroups. Estimates were obtained from IPTW-weighted logistic regression fitted separately within each subgroup. The vertical line indicates the null value (OR = 1).

Fig. 3.

Exploratory subgroup analyses of the association between corticosteroids plus antiviral therapy and the composite outcome (death or mechanical ventilation). Forest plot of odds ratios (ORs) and 95% confidence intervals (CIs) for the composite outcome in the IPTW-weighted cohort, by subgroup. ORs were estimated from IPTW-weighted logistic regression within each subgroup (stabilized weights, 1%–99% truncation). The vertical line indicates OR = 1.

3.5. Assessment of propensity score weighting

Propensity score diagnostics demonstrated adequate overlap of propensity score distributions between treatment groups, indicating sufficient common support for weighting. The distribution of stabilized IPTW weights after truncation at the 1st and 99th percentiles showed no evidence of extreme weight instability (Fig. 4).

Fig. 4.

Assessment of propensity score overlap and distribution of IPTW weights. (A) Kernel density of the propensity score (PS) by treatment group (corticosteroids alone vs corticosteroids plus antiviral). Overlap between the two distributions indicates adequate common support for weighting. (B) Distribution of stabilized inverse probability of treatment weighting (IPTW) weights after truncation at the 1st and 99th percentiles. The propensity score model included age, sex, center, comorbidities, and disease severity.

3.6. Sensitivity analysis

A sensitivity analysis incorporating baseline laboratory variables into the propensity score model was performed in the complete-case population with available laboratory data. After IPTW adjustment, baseline covariates remained well balanced between groups, and the direction and magnitude of treatment effect estimates were consistent with those observed in the primary analysis (Table S1).

4. Discussion

In this multicenter real-world retrospective study of elderly hospitalized patients with COVID-19 during the Omicron period who received systemic corticosteroid therapy, corticosteroid–antiviral combination therapy was associated with significantly lower in-hospital mortality compared with corticosteroid monotherapy, whereas no significant difference was observed in the risk of mechanical ventilation. These findings suggest that adding antiviral therapy to corticosteroids may provide additional survival benefit in this high-risk treatment population.

Our findings are broadly consistent with previous hospital-based studies suggesting that antiviral–corticosteroid combination therapy may improve outcomes in patients with COVID-19.^16,17 Evidence from nirmatrelvir/ritonavir studies in high-risk or hospitalized populations, including real-world cohorts during the Omicron era, also supports the potential benefit of antiviral therapy in patients at elevated risk for progression ^10,18, 19, 20, 21. This association is biologically plausible, as corticosteroids primarily attenuate the dysregulated host inflammatory response, whereas antiviral agents inhibit viral replication; the combination may therefore provide complementary therapeutic effects.

In contrast, we did not detect a significant reduction in mechanical ventilation. This differs from some previous reports in which benefit was observed mainly for composite disease-progression outcomes rather than for mechanical ventilation alone.¹⁸ The discrepancy may therefore partly reflect differences in endpoint definition and clinical decision-making regarding invasive ventilation. In elderly hospitalized patients, the decision to initiate mechanical ventilation is influenced by frailty, comorbidities, ceiling-of-care decisions, patient and family preferences, overall prognosis, and local clinical practice. In addition, the antiviral regimen in our cohort predominantly consisted of nirmatrelvir/ritonavir, whereas some earlier studies were based mainly on remdesivir-containing regimens.¹⁷ Differences in patient characteristics, baseline severity, oxygen requirement, treatment setting, and timing of antiviral initiation may also have contributed to the heterogeneous findings across studies. Residual confounding cannot be fully excluded despite IPTW adjustment.

This study has several strengths. First, multicenter real-world evidence from China directly comparing corticosteroid monotherapy with corticosteroid–antiviral combination therapy in elderly hospitalized patients with COVID-19 remains limited, particularly during the Omicron era. Second, our study focused on an elderly population, which is clinically important because older adults remain at particularly high risk for adverse outcomes. Third, rather than evaluating antiviral therapy alone, we examined its incremental value when added to corticosteroid therapy, which more closely reflects real-world treatment decision-making in hospitalized patients receiving corticosteroids. Finally, the use of IPTW improved comparability between treatment groups and strengthened the robustness of the main findings in an observational setting.

Several limitations should be acknowledged. First, this was a retrospective observational study, and residual confounding from unmeasured variables cannot be fully excluded despite IPTW adjustment. Second, the interval from symptom onset to treatment initiation, the exact timing of antiviral initiation, and the detailed dose and duration of corticosteroid therapy were not fully standardized or systematically captured in this retrospective dataset, which may have influenced treatment effects. Therefore, treatment timing could not be reliably adjusted for in the statistical models, and standardized time-to-treatment analyses were not possible. Third, treatment-related adverse events, such as secondary infection, hyperglycemia, and gastrointestinal bleeding, were not systematically collected and therefore could not be compared between groups. Fourth, exploratory subgroup analyses did not identify significant heterogeneity of treatment effect, but these analyses were limited by relatively small numbers of events within some strata and should be interpreted cautiously. Finally, long-term outcomes after hospital discharge were not available, so the potential impact of treatment strategy on post-discharge recovery and longer-term prognosis could not be assessed.

Overall, in elderly hospitalized patients with COVID-19 during the Omicron period who received systemic corticosteroid therapy, corticosteroid–antiviral combination therapy was associated with lower in-hospital mortality than corticosteroid monotherapy, although no significant difference was observed for mechanical ventilation. Larger prospective studies are needed to confirm these findings and to clarify which patient subgroups may derive the greatest benefit from combination therapy.

5. Conclusions

In this multicenter real-world cohort of elderly patients hospitalized with COVID-19 during the Omicron period who received systemic corticosteroid therapy, corticosteroids combined with antiviral therapy were associated with lower in-hospital mortality compared with corticosteroid monotherapy, while no significant difference was observed in the risk of mechanical ventilation. These findings suggest that combination therapy targeting both viral replication and inflammatory responses may improve survival in this high-risk treatment population.

CRediT authorship contribution statement

Shenglin Xu: Writing – review & editing, Writing – original draft, Visualization, Formal analysis. Zhaohui Huang: Writing – original draft, Investigation, Formal analysis, Data curation. Xuejing Zou: Resources, Project administration. Guiping Li: Project administration. Richeng Mao: Supervision, Data curation. Li Liu: Supervision, Funding acquisition, Conceptualization. Yue Guo: Writing – review & editing, Supervision, Methodology. Xiaoyong Zhang: Writing – review & editing, Supervision, Resources, Project administration, Funding acquisition, Conceptualization.

Consent for publication

All authors approved the manuscript for publication.

Funding

This study was supported by grants from the National Key Research and Development Program of China (Grant No. 2023YFC2308504).

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A. Supplementary data

The following are the Supplementary data to this article:

Data availability

Data will be made available on request.