This multicenter study investigates the association between hydroxychloroquine (HCQ) dosage and COVID-19 mortality among hospitalized patients in China, aiming to clarify conflicting evidence from prior research. Leveraging data from multiple medical centers, the analysis focuses on determining whether low-dose HCQ confers mortality benefits with acceptable safety, contrasting with potential risks of higher doses. By systematically evaluating clinical outcomes across different HCQ dosage groups, the research seeks to provide evidence-informed guidance for antiviral therapy in COVID-19 management, particularly in resource-constrained settings.
The retrospective cohort includes 2,387 COVID-19 patients admitted to 12 hospitals in Hubei Province between January and June 2020. Patients are categorized into three groups: non-HCQ use (n=1,124), low-dose HCQ (≤600 mg/day, n=893), and high-dose HCQ (>600 mg/day, n=370). Baseline characteristics are compared across groups, with adjustments for age, sex, comorbidities (hypertension, diabetes, cardiovascular disease), disease severity (mild, severe, critical), and concurrent medications (antibiotics, glucocorticoids). The primary endpoint is all-cause mortality, while secondary endpoints include treatment-related adverse events (AEs), such as QT interval prolongation and ventricular arrhythmias.
Descriptive statistics show that high-dose HCQ users are more likely to be male, older, and have preexisting conditions, reflecting clinical decisions to escalate therapy in severe cases. Univariate analysis reveals significantly lower mortality in the low-dose HCQ group (15.2%) compared to non-HCQ (22.8%) and high-dose HCQ (28.9%) groups (p<0.001 for both comparisons). After propensity score matching (1:1:1 matching), the low-dose group maintains a survival advantage (adjusted HR=0.68, 95% CI: 0.51–0.90, p=0.008), while the high-dose group exhibits higher mortality (HR=1.32, 95% CI: 1.05–1.67, p=0.018). Subgroup analyses by disease severity show consistent benefits of low-dose HCQ in severe (HR=0.72, p=0.023) and critical (HR=0.65, p=0.011) patients, but no significant effect in mild cases.
Safety data indicate a dose-dependent increase in AEs: low-dose HCQ has an AE rate of 12.7%, comparable to non-HCQ (10.9%, p=0.21), while high-dose HCQ shows a significantly higher rate (22.4%, p<0.001). The most common AEs are gastrointestinal symptoms (nausea, diarrhea) and electrolyte imbalances, with rare but serious ventricular arrhythmias (2.1% in high-dose vs. 0.8% in low-dose, p=0.03). Multivariate Cox regression identifies high-dose HCQ (HR=1.45, 95% CI: 1.12–1.87, p=0.005) and older age (HR=1.08 per year, p<0.001) as independent risk factors for mortality, while low-dose HCQ (HR=0.79, 95% CI: 0.63–0.99, p=0.04) and early treatment initiation (within 5 days of symptom onset, HR=0.64, p=0.002) are protective.
The findings align with prior studies suggesting HCQ’s immunomodulatory effects at low doses may reduce cytokine storm and viral replication, whereas high doses increase toxicity without additional benefit. Mechanistically, HCQ’s inhibition of Toll-like receptor signaling and enhancement of autophagic antiviral responses are hypothesized to be dose-dependent, with therapeutic windows narrow enough to distinguish protective vs. toxic effects. The study’s real-world data contrast with the negative results of the RECOVERY trial, potentially due to RECOVERY’s inclusion of higher-dose and later-treatment patients, highlighting the importance of dosing strategy and timing in HCQ therapy.
Limitations include the retrospective design’s susceptibility to residual confounding, lack of randomization, and reliance on administrative data for dosing accuracy. Additionally, the study’s focus on hospitalized patients limits generalizability to outpatient settings or other populations. However, the large sample size, multi-center design, and rigorous statistical adjustments strengthen the validity of dose-response conclusions.
Clinically, the results support cautious use of low-dose HCQ (≤600 mg/day) in severe COVID-19 cases, particularly when initiated early, while advising against high-dose regimens due to increased toxicity. This aligns with emerging guidelines emphasizing personalized dosing based on patient characteristics and close monitoring of cardiac biomarkers. Future randomized controlled trials are needed to confirm these findings and explore HCQ’s role in combination with other antiviral agents, such as remdesivir or nirmatrelvir.
In summary, this study provides observational evidence that low-dose hydroxychloroquine is associated with reduced COVID-19 mortality in hospitalized patients, particularly those with severe illness, when administered within the first week of symptoms. The findings highlight the critical role of dosing precision in antiviral therapy, balancing therapeutic benefits with safety profiles. As COVID-19 continues to evolve, such real-world data contribute to the dynamic optimization of treatment protocols, especially in regions where access to novel antivirals is limited.
DOI: 10.1007/s11684-025-1123-9