Table 3. Adjusted Pooled Odds of COVID-19–Associated Hospitalization or Death Among Patients in the 3 Registriesa.
| Variable | OR (95% CI) | P value |
|---|---|---|
| Exposure treatment regimenb | ||
| TNF inhibitor monotherapy | 1 [Reference] | NA |
| TNF inhibitor plus methotrexate | 1.18 (0.85-1.63) | .33 |
| TNF inhibitor plus azathioprine/6-mercaptopurine | 1.74 (1.17-2.58) | .006 |
| Methotrexate monotherapy | 2.00 (1.57-2.56) | <.001 |
| Azathioprine/6-mercaptopurine monotherapy | 1.84 (1.30-2.61) | .001 |
| Jak inhibitor monotherapy | 1.82 (1.21-2.73) | .004 |
| Concomitant medication | ||
| Sulfasalazine | 1.62 (1.13-2.34) | .009 |
| Leflunomide | 1.89 (1.20-2.99) | .006 |
| Oral budesonide | 2.86 (1.20-6.84) | .02 |
| Daily prednisone-equivalent dose per 1 mg increase | 1.07 (1.05-1.08) | <.001 |
| Demographic characteristic | ||
| Female sex | 0.79 (0.66-0.96) | .02 |
| Age per year | 1.04 (1.04-1.05) | <.001 |
| Obesity (BMI ≥30) | 1.39 (1.10-1.75) | .005 |
| Current smoking | 0.77 (0.51-1.17) | .21 |
| Disease activity | ||
| Active | 1.27 (1.04-1.55) | .02 |
| Comorbidities | ||
| Interstitial lung disease | 1.81 (1.12-2.95) | .02 |
| Obstructive lung disease | 2.34 (1.69-3.24) | <.001 |
| Cardiovascular disease | 1.58 (1.13-2.21) | .007 |
| Diabetes | 1.54 (1.16-2.05) | .003 |
| Hypertension | 1.19 (0.95-1.50) | .12 |
| Kidney disease | 3.10 (1.70-5.66) | <.001 |
| Cancer | 1.16 (0.65-2.07) | .61 |
Abbreviations: BMI, body mass index (calculated as weight in kilograms divided by height in meters squared); Jak, Janus kinase; NA, not applicable; OR, odds ratio; TNF, tumor necrosis factor.
All 6077 patients from the COVID-19 Global Rheumatology Alliance (GRA); the Psoriasis Patient Registry for Outcomes, Therapy and Epidemiology of COVID-19 Infection (PsoProtect); and the Secure Epidemiology of Coronavirus Under Research Exclusion for Inflammatory Bowel Disease (SECURE-IBD) registries were included.
Subcategories are mutually exclusive. Odds ratios were obtained using hierarchical multivariable mixed-effects logistic regression analysis with registry and calendar month random effects nested within country. Model was adjusted for all variables shown.