• tumor necrosis factors
• methotrexate
• biological therapy
• COVID-19
• epidemiology

Susceptibility and tolerance to COVID-19 of patients with rheumatic disorders remains poorly understood. A recent meta-analysis did not demonstrate any considerably worse outcomes.1 Sufferers from inflammatory rheumatic disorders are, however, known to be more prone to infections than the general population and this risk is increased by targeted biologic therapy.2 3 Therefore, we were interested in examining the risk of admission and respiratory failure in patients with rheumatic disorders with COVID-19 being treated with targeted synthetic or biologic disease-modifying anti-rheumatic drugs (ts/bDMARDs), comparing them to matched comparators and methotrexate (MTX) users within Iceland. Unique conditions exist in Iceland for this study, as the island nation is naturally isolated and performed extensive screening, tracing and systematic registration of all PCR-confirmed cases.4 5 All diagnosed individuals received regular follow-up by a COVID-19 outpatient clinic.6

ICEBIO is a nationwide registry of patients with inflammatory arthritis treated with ts/bDMARDs. We included all patients in ICEBIO undergoing treatment at the start of the domestic outbreak. From the Icelandic Medicine Database we extracted all MTX prescriptions filled in the 9 months before Iceland’s first recorded case of COVID-19. Each individual from the ICEBIO and MTX groups was randomly matched with up to ten controls based on age, sex and geographic location. Individuals in ICEBIO or with MTX prescriptions from haematologists and oncologists were excluded from the MTX group, although their comparator group remained unaltered. The Icelandic Directorate of Health provided data on all PCR tests and hospital admissions in our study population. Data were extracted on 3 June 2020, when the first domestic outbreak ended: 1796 individuals had been diagnosed with COVID-19, with two active cases remaining. At that time, 61 639 tests had been administered in a nation of roughly 360 000 people, and the Directorate of Health reports a successful infection tracing rate of over 95%.4

We identified 1438 individuals from ICEBIO, 13 815 ICEBIO comparators, 1746 individuals with an MTX prescription and 22 962 MTX comparators, see table 1. The relative risk (RR) for the ICEBIO group to undergo testing was 1.35 (1.23–1.48; p<0.001), compared with their comparators and the RR for the MTX group at 1.05 (0.96–1.15; p=0.28) compared with theirs.

Nine from ICEBIO, eighty-four ICEBIO comparators, five MTX treated and one hundred and thirty-four MTX comparators were SARS-CoV-2 positive. All infected patients from ICEBIO had received tumour necrosis factor inhibitors (online supplemental tables S1 and S2). Two of three hospitalised patients from ICEBIO, three of three ICEBIO comparators, one of one from the MTX group and ten of thirteen hospitalised MTX comparators received oxygen supplementation. One of three admitted patients from the ICEBIO comparators and two of thirteen MTX comparators received mechanical ventilation (table 1). The RR for infected patients from ICEBIO to be admitted was 9.33 (2.20–39.6; p<0.001) and 6.22 (1.19–32.46; p=0.02) to be admitted with hypoxia. The RR for hypoxia following admission was 0.67 (0.30–1.48) for patients from ICEBIO and 0.77 (0.57–1.4) for patients taking MTX. The mean length of admission for the patient from ICEBIO was 4.7±3.6 days, while it was 20.2±12.7 days for their comparators (p=0.16). As no patients with rheumatic disorders in any group required mechanical ventilation, neither OR nor RR can be calculated for that outcome.

We found that patients with COVID-19 with rheumatologic disorders on bDMARDs are at a higher risk of hospitalisation than matched comparators. This might be explained by a lower threshold for admitting patients on biologics, as hospitalised rheumatology patients on bDMARDs fared numerically better, although the small numbers prevent meaningful statistical analysis. Further studies in larger populations are needed to better quantify the risk and severity of COVID-19 in patients with rheumatic disorders treated with bDMARDs.