The association between immunosuppressants use and COVID-19 adverse outcomes: National COVID-19 cohort in South Korea

Hyun Joon Shin, Ronald Chow, Hyerim Noh, Jongseong Lee, Jihui Lee, Young‐Geun ChoimedRxiv 2021 · DOI

Abstract

There has been significant clinical uncertainty regarding how pre-existing immunosuppression—particularly the use of corticosteroids—affects the trajectory of COVID-19. This population-based retrospective cohort study utilized a nationwide healthcare claims database from South Korea to investigate the relationship between pre-hospitalization immunosuppressant use and patient outcomes. Analyzing 4,349 hospitalized COVID-19 patients aged 40 and older, the study employed Inverse Probability of Treatment Weighting (IPTW)-adjusted logistic regression to compare users and non-users. The primary outcome was a composite endpoint of all-cause death, intensive care unit (ICU) admission, and mechanical ventilation. The results indicate that immunosuppressant users faced significantly higher odds of severe outcomes (IPTW OR 1.32; 95% CI: 1.06–1.63), with corticosteroid users specifically showing similar increased risk. These findings support clinical guidelines prioritizing immunocompromised individuals for booster vaccinations and intensive monitoring.

Introduction

Early in the COVID-19 pandemic, the medical community faced a paradox regarding immunosuppressive therapy. While corticosteroids eventually became a standard of care to mitigate the lethal "cytokine storms" seen in late-stage severe infections, the impact of maintenance immunosuppression for chronic conditions remained unclear. Does a dampened immune system protect a patient from hyper-inflammation, or does it facilitate viral replication and hinder the initial defense?

Previous studies into this question often suffered from small sample sizes or specific geographic biases. This research sought to provide a definitive, large-scale analysis using the comprehensive healthcare data available in South Korea to determine if pre-existing immunosuppressant use correlates with worse clinical outcomes.

Methods

Study Design and Population

This study was a population-based retrospective cohort study using the #OpenData4COVID19 project, a de-identified dataset released by the Ministry of Health and Welfare of Korea. The data originated from the Health Insurance Review and Assessment Service (HIRA) database, which covers approximately 98% of the South Korean population.

The researchers initially screened 234,427 individuals tested for SARS-CoV-2. The final analysis focused on 4,349 confirmed COVID-19 patients who were hospitalized and aged 40 years or older. By linking these records to three years of prior claims data, the team established a detailed medical history for each patient.

Exposure and Outcome Measures

  • Exposure: "Users" were defined as patients with at least one prescription for immunosuppressants (including corticosteroids, TNF antagonists, etc.) within 120 days prior to hospital admission.
  • Primary Outcome: A composite endpoint consisting of all-cause death, ICU admission, and the requirement for mechanical ventilation.
  • Statistical Analysis: To account for confounding variables in observational data, the study utilized Inverse Probability of Treatment Weighting (IPTW). This balanced the user and non-user groups across 20 different comorbidities (e.g., diabetes, hypertension, chronic obstructive pulmonary disease) as well as demographic factors like age and sex.

Results

Of the 4,349 patients in the cohort, 1,356 (31%) were classified as immunosuppressant users. Within this group, corticosteroids were the dominant medication, accounting for 98.8% of the prescriptions.

Primary and Secondary Outcomes

  • Composite Outcome: Patients using immunosuppressants had a 32% increase in the odds of reaching the composite endpoint (IPTW OR 1.32; 95% CI: 1.06–1.63).
  • Mortality: When analyzed as a standalone metric, the risk was even more pronounced. Immunosuppressant users had an IPTW OR of 1.63 (95% CI: 1.21–2.26) for all-cause death.
  • Corticosteroid Specificity: The results for those specifically on corticosteroids mirrored the broader group, with an OR of 1.33 for the primary composite outcome.

Sensitivity analyses—including varying the exposure window to 90 days and using Propensity Score matching—confirmed the robustness of these findings.

Discussion

The Timing Paradox

The study’s findings highlight a critical distinction in the timing of steroid use. While clinical trials like RECOVERY demonstrated that dexamethasone reduces mortality when administered during severe COVID-19, this study suggests that being on these drugs at the time of infection is detrimental. Pre-existing immunosuppression likely impairs the body's initial innate and adaptive immune responses, potentially leading to higher viral loads and a more difficult path to clearance.

Breadth of the Dataset

A unique strength of this study is the nature of the South Korean healthcare response during the study period. Early in the pandemic, South Korea hospitalized nearly all confirmed cases, including those who were asymptomatic or mildly symptomatic. This allows for a more comprehensive view of the disease spectrum than studies from regions that only hospitalized the critically ill, reducing the likelihood of selection bias.

Limitations

As a claims-based retrospective study, there are inherent limitations. The researchers lacked access to granular clinical data such as Body Mass Index (BMI), baseline laboratory values (e.g., C-reactive protein or viral load), and the specific severity of symptoms at the exact moment of diagnosis. Furthermore, while IPTW is a powerful tool for balancing cohorts, the observational nature of the study means it can identify associations but cannot definitively prove a causal link.

Conclusion

This nationwide cohort study provides strong evidence that patients on maintenance immunosuppressants, particularly corticosteroids, are at a significantly higher risk for severe COVID-19 outcomes and death. These results align with and support the latest public health guidelines, such as those from the CDC, which categorize immunocompromised individuals as high-risk. The findings underscore the importance of prioritizing these populations for booster vaccinations and suggest that clinicians should maintain a high index of suspicion for clinical deterioration in patients with a history of pre-hospitalization immunosuppressant use.

Summary updated

Abstract

Summary Purpose There is uncertainty of the effect of immunosuppression, including corticosteroids, before COVID-19 infection on COVID-19 outcomes. The aim of this study was to investigate the relationship between prehospitalization immunosuppressants use (exposure), and COVID-19 patient outcomes. Methods We conducted a population-based retrospective cohort study using a nationwide healthcare claims database of South Korea as of May 15, 2020. Confirmed COVID-19 infection in hospitalized individuals aged 40 years or older were included for analysis. We defined exposure variable by using inpatient and outpatient prescription records of immunosuppressants from the database. Our primary outcome was a composite endpoint of all-cause death, intensive care unit (ICU) admission, and mechanical ventilation use. Inverse probability of treatment weighting (IPTW)-adjusted logistic regression analyses were used, to estimate odds ratio (OR) and 95% confidence intervals, comparing immunosuppressants users and non-users. Results We identified 4,349 patients, for which 1,356 were immunosuppressants users and 2,903 were non-users. Patients who used immunosuppressants were at increased odds of the primary outcome of all-cause death, ICU admission and mechanical ventilation use (IPTW OR 1.32; 95% CI: 1.06 – 1.63). Patients who used corticosteroids were at increased odds of the primary outcome (IPTW OR 1.33; 95% CI: 1.07 – 1.64). Conclusion We support the latest guidelines from the CDC, that people on immunosuppressants are at high risk of severe COVID-19 and immunocompromised people may need booster COVID-19 vaccinations. Funding YGC’s work was partially supported by 2020R1G1A1A01006229 awarded by the National Research Foundation of Korea.
BibTeX
@misc{shin2021association,
  author = {Hyun Joon Shin and Ronald Chow and Hyerim Noh and Jongseong Lee and Jihui Lee and Young‐Geun Choi},
  title = {{The association between immunosuppressants use and COVID-19 adverse outcomes: National COVID-19 cohort in South Korea}},
  year = {2021},
  doi = {10.1101/2021.08.17.21262183},
  howpublished = {medRxiv},
}
Figure