2196: Risk Factors for Severe COVID-19 Outcomes: A Study of Immune-mediated Inflammatory Diseases, Immunomodulatory Medications, Vaccination Status, and Comorbidities in a Large US Healthcare System

Philip J Mease¹, Qi Wei², Michael Chiorean¹, Lulu Iles-Shih¹, Wanessa Matos², Andrew Baumgartner², Sevda Molani², Yeon Mi Hwang², Basazin Belhu² and Jennifer Hadlock², ¹Swedish Medical Center/Providence St. Joseph Health, Seattle, WA, ²Institute for Systems Biology, Seattle, WA

Background/Purpose: Reported infection rates and severe outcomes of COVID-19 in patients with immune-mediated inflammatory diseases (IMIDs) and their therapy, have varied depending on the nature and size of the study population, and the presence or absence of control populations. Having a more reliable analysis of larger populations, with consideration of vaccination status and comorbidities is essential for current and future pandemics.

Methods: All patients (pts) testing positive by SARS-CoV-2 PCR and NAAT test were identified from 51 hospitals and 1,085 clinics in Providence St. Joseph Health, which serves seven US states. IMID pts, subsetted into those typically seen by a rheumatologist and those by other specialists were identified. Outcomes were hospitalization and death within 30 days of the first COVID infection. Two time periods were analyzed: 03/01/2020 – 12/25/2021: pre-Omicron, 12/26/2021 – 04/13/2022: Omicron. Immunomodulatory medications were observed for three months prior to the index date. Risk factors, including IMIDs, medications, demographics, comorbidities, and vaccination status were analyzed using multivariable logistic regression and gradient boosting (XGBoost).

Results: In the pre-Omicron period, 1,476,322 people were tested with 11% positive. Of these, 2.5% were hospitalized and 3.2% died. 78,061 IMID pts were tested with 7.9% positive, and of these 3% were hospitalized and 4.5% died, higher rates than the overall population. The rheumatologic IMIDs had greater hospitalization and mortality rates compared to other IMIDs. In the Omicron period, the positive rate increased (21.9%) but the percentage hospitalized (0.8%) and dying (1.8%) decreased, partly due to increased vaccination status and possibly decreased viral virulence. In both time periods, pts with IMIDs had an increased rate of hospitalization and death compared to non-IMID pts (Tab 1). XGBoost models achieved AUROC 0.74 - 0.91 on the test set. Age and multiple comorbidities were strongly associated with an increased risk of hospitalization and death, whilst vaccination and booster status were strongly associated with a decreased risk of hospitalization and death. Pts with IMIDs overall had higher percentages of hospitalization and death than those without, likely associated with age and comorbidities. However, few IMIDs were associated with severe outcomes, and those had low feature importance (Figs 1, 2). Spondyloarthritis and Sjögren's were less associated with death, as was the use of TNFi. The favorable or unfavorable association of specific IMIDs and treatments with outcomes varied by time period and analytic method (Figs 1, 2).

Conclusion: Pts with IMIDs, especially rheumatologic, experienced more severe outcomes than non-IMID pts. Most immunomodulatory medications were not associated with increased risk. Some IMIDs and treatments were associated with decreased risk of severe outcomes, such as spondyloarthritis and Sjögren's, as well as the use of TNFi. Certain drug classes may be weakly associated with increased risk but due to the inconsistency of this signal with different analytic methods and time periods, these observations need further research.
Table 1. COVID-19 test results, hospitalization, and mortality. IMID: rheumatoid arthritis, spondyloarthritis, systemic lupus erythematosus, psoriatic arthritis, systemic sclerosis, vasculitis, sarcoidosis, antiphospholipid syndrome, Sjögren's, inflammatory bowel disease, multiple sclerosis, psoriasis; Rheumatologic IMIDs: rheumatoid arthritis, spondyloarthritis, systemic lupus erythematosus, psoriatic arthritis, systemic sclerosis, vasculitis, sarcoidosis, antiphospholipid syndrome, Sjögren's. Other IMIDs: inflammatory bowel disease, multiple sclerosis, psoriasis.

Figure 1: Log adjusted odds ratio (OR). Selected risk factors for hospitalization and mortality in COVID-19 positive patients, using multivariable logistic regression (LR). The color of dots represents P-value from under-sampling. position and size of dots represent log-adjusted odds ratio and feature importance from over-sampling. The error bar represents the 95% confidence interval. Factors with fewer than 20 observations were excluded. NE: not enough samples to calculate the p-value after under-sampling.

Figure 2: (A) Feature importance of top 25 features of the all-age model of COVID-19 hospitalization within 30 days at two-time intervals. (B) Feature importance of top 25 features of the all-age model of COVID-19 death within 30 days at two-time intervals. Gradient boosting decision tree’s feature importance and influence of higher and lower values of the risk factors on the all-age group population outcome. The left side of each plot is associated with a reduced risk of hospitalization or death and the right side of each plot is associated with an increased risk of hospitalization or death. Red dots represent patients with higher values for a variable, and blue dots are patients with lower values. Nominal classes are binary [0, 1]. For sex, male is 1 (red).
Disclosures: P. Mease, AbbVie, Amgen, Janssen, Novartis, Pfizer Inc, UCB, Sun Pharma, Eli Lilly, Bristol-Myers Squibb(BMS), Celgene, Genentech; Q. Wei, Pfizer, Swedish Medical Foundation, Novartis; M. Chiorean, Pfizer, Bristol-Myers Squibb(BMS), Takeda, AbbVie/Abbott, Janssen, Medtronic, Eli Lilly, Arena, Novartis, Swedish Medical Foundation; L. Iles-Shih, Pfizer, Novartis, Swedish Medical Foundation; W. Matos, Pfizer, Novartis, Swedish Medical Foundation; A. Baumgartner, None; S. Molani, Pfizer, Novartis, Swedish Medical Foundation; Y. Hwang, None; B. Belhu, Pfizer, Novartis, Swedish Medical Foundation; J. Hadlock, Pfizer, Novartis, Swedish Medical Foundation.