ABSTRACT
Background: The risk of acquiring COVID-19, and the severity of illness if acquired, in the context of immune-mediated infammatory diseases (IMIDs) and their therapy, remains incompletely understood. Reported infection rates and outcomes have varied depending on the IMIDs being studied, the nature and size of the study population, and the presence or absence of appropriate control populations. Having more reliable analysis on larger populations is essential for current and future pandemics. Objectives: Health records from one of the largest health systems in the US are analyzed to determine whether specifc IMIDs, including common rheumatologic conditions and specifc immunomodulatory drugs, are associated with certain COVID-19 outcomes, using multivariate models that include common chronic comorbidities. Methods: Patients (pts) with and without IMIDs who were tested for SARS-CoV-2 antigen (n=1,101,431) were identifed from the EHR from Providence St. Joseph Health, which serves much of the western US. Immunomodulatory drug therapy was defned as use within three months prior to the frst test. Multivariate logistic regression (LR) was applied with machine learning metrics (feature importance, p-value) reported on an 80% training set and AUROC reported on 20% test set. Results: Rates for positive COVID-19 tests, invasive mechanical ventilation (IMV) and mortality were not greater in the IMID than non-IMID population, whilst hospitalization was similar (Table 1). Importance and statistical signifcance of selected factors are shown in (Figure 1). The most important risk factors for hos-pitalization were age and heart failure. Heart failure was the most important risk factor for IMV, and age for increased mortality. Diabetes showed weak associations with these three outcomes. Spondyloarthritis was weakly associated with decreased hospitalization, IMV, and death. The use of conventional synthetic disease-modifying antirheumatic drugs (csDMARDS) and corticosteroids (CS) showed a weak association with hospitalization, and rituximab (RTX) showed a weak association with increased mortality. Limitations include lack of vaccination status and IMID disease severity/fare status. Testing was not universal. Conclusion: This analysis of COVID+ patients (n=1,101,431) from a large US health care system analyzes outcomes of patients with and without IMIDs;the majority were rheumatologic IMIDs. Patients with IMIDs had a similar rate of hospitalization, IMV, and death as those without IMIDs. The strongest associations with COVID-19 severity included heart failure and age. Spondyloarthritis was weakly associated with favorable outcomes whilst other conditions, including rheumatologic, were not worse than those of non-IMID patients. csDMARDs and corticosteroids were weakly associated with hospitalization and RTX with increased mortality. Other therapies were not associated with severe adverse outcomes.
ABSTRACT
Introduction: Patients with IBD receiving immunosuppressive (IS) therapy may be at higher risk of infectious complications. The impact of IS drugs on COVID-19 outcomes in patients with IBD and other immune-mediated inflammatory disorders (IMIDs) is unclear. We aimed to determine COVID-19 outcomes in a large population database from a single healthcare system. Methods: Adult patients with IBD and other IMIDs were identified using SNOMED-CT codes from the EMR of the Providence Health System. COVID-19 diagnoses were identified based on SARSCoV-2 positive PCR results. COVID-19 incidence, hospitalization, ICU admission and death were analyzed among IBD, other IMID patients and a control group without IMIDs. Risk factors for severe COVID-19 outcomes were estimated using logistic regression with demographics, comorbidities and IMID drugs including corticosteroids, biologics and JAK inhibitors as independent variables. We considered defined drug exposure as having a drug prescribed or administered at least once within the year prior to the first positive SARS-CoV-2 positive test result. Results: We identified 58,612 adult patients with IBD (32,043 UC and 26,569 CD), 130,799 with other IMIDs, and 6,0±6,081 controls without IMID. Among these, there were 59,647 COVID-19 cases (0.95%) including 427 (0.73%) among IBD patients. Of these 80 (18.7%) were hospitalized, 3 (0.7%) required ventilatory support and 10 died (2.3%). COVID-19 incidence was similar among IBD, other IMID and non-IMID controls. Risk factors for severe COVID-19 outcomes are shown in the Table 1. Among IBD patients, only age (aOR 1.04, p=0.0001) was associated with an increased risk of hospitalization whereas steroid, IMM or advanced therapeutic use was not. Age was also a risk factor for hospitalization among other IMID patients whereas the use of immunotherapeutic drugs appeared to be protective. Conclusion: Patients with IBD and other IMIDs do not appear to be at increased risk of COVID-19 infections or related serious complications. Age is the only variable associated with an increased risk factor for hospitalization among patients with IBD and other immune-mediated conditions. Further studies on this topic are warranted.