Introduction
At the onset of the COVID-19 pandemic, public health authorities frequently advised individuals with autoimmune diseases (AD) to "shield" or practice strict isolation. This guidance was based on the theoretical risk that both their underlying conditions and the use of immunosuppressive therapies would lead to poorer outcomes. However, empirical evidence to support these recommendations was initially limited to small, localized studies.
To address this knowledge gap, researchers conducted a massive multinational distributed network cohort analysis. The study aimed to characterize the socio-demographics, comorbidities, and 30-day outcomes of patients with prevalent autoimmune conditions who contracted COVID-19, while providing a critical comparison against historical outcomes for seasonal influenza.
Methodology
The study was conducted as part of the CHARYBDIS (Characterizing Health Associated Risks, and Your Baseline Disease In SARS-COV-2) project. To ensure patient privacy while maintaining large-scale data integrity, the researchers utilized a federated analysis approach via the OMOP Common Data Model (CDM). This allowed analytical code to be run locally across various global databases, with only aggregated results shared for synthesis.
Data Sources
The analysis leveraged six major data sources across the United States, Spain, and South Korea, representing a diverse cross-section of healthcare environments:
- CUIMC: Columbia University Irving Medical Center (NYC, US)
- Optum: De-identified Electronic Health Records (US)
- VA-OMOP: Department of Veterans Affairs (US)
- SIDIAP-H: Primary care records linked to hospital data (Spain)
- IQVIA Open Claims: US-based claims data
- HIRA: Health Insurance and Review Assessment (South Korea)
Cohort Definitions
The study identified 133,589 patients with AD diagnosed with COVID-19 and 48,418 who were hospitalized between January and June 2020. These cohorts were compared to a historical group of patients with AD hospitalized with seasonal influenza between September 2017 and April 2018.
The Vulnerability Profile
The study revealed a consistent profile for AD patients who required hospitalization for COVID-19. The majority were women, aged 50 years or older, and presented with significant pre-existing comorbidities.
Patients hospitalized with COVID-19 showed a higher prevalence of hypertension, chronic kidney disease, heart disease, and Type 2 diabetes compared to those who were diagnosed but did not require hospitalization. While psoriasis and rheumatoid arthritis were the most common autoimmune conditions in the general cohort, Type 1 Diabetes was disproportionately prevalent among those hospitalized in the US-based databases.
COVID-19 vs. Seasonal Influenza
One of the most significant findings of this research is the stark contrast between COVID-19 and seasonal influenza for this patient population. The data provides a quantitative "reality check" against the early-pandemic narrative that COVID-19 was comparable to the flu.
- Mortality Gap: The 30-day mortality rate for hospitalized COVID-19 patients with AD ranged from 6.3% to 24.6% across different regions. In comparison, the mortality rate for the influenza cohort was significantly lower, ranging from 2.2% to 4.3%.
- Hospitalization Rates: Patients with AD diagnosed with COVID-19 were more likely to require hospitalization (up to 36.7%) than those diagnosed with influenza (up to 23.6%).
Clinical Complications
Patients with autoimmune diseases hospitalized with COVID-19 faced severe clinical complications within 30 days of admission:
- Respiratory Failure: Acute Respiratory Distress Syndrome (ARDS) occurred in 2.1% to 42.8% of cases, while pneumonia was diagnosed in 12.6% to 53.2%.
- Renal Impact: Acute Kidney Injury (AKI) emerged as a major complication, particularly in US databases, where it affected up to 31.1% of patients.
Discussion and Limitations
The study noted significant regional variance in mortality, with South Korea reporting 6.3% compared to nearly 25% in parts of the US and Spain. These differences likely reflect variations in healthcare system strain, testing availability, and regional demographics during the first wave of the pandemic.
Study Limitations
- Testing Bias: Early in the pandemic, limited testing availability likely meant that many mild cases were not captured, potentially skewing the data toward more severe outcomes.
- Data Coding: In US insurance claims, Type 2 Diabetes is occasionally miscoded as Type 1, which may have influenced the reported prevalence of specific autoimmune conditions.
- Mortality Underestimation: EHR-based databases may fail to capture deaths occurring after hospital discharge, suggesting that the reported mortality rates could be underestimates.
Conclusion
This multinational analysis confirms that patients with autoimmune diseases are at high risk for severe complications and death from COVID-19. The findings underscore that for this vulnerable population, COVID-19 is substantially more lethal and more likely to cause catastrophic respiratory and renal failure than seasonal influenza. Future research is needed to determine if specific immunosuppressive medications further exacerbate these risks.
Summary updated