Abstract
Background
During the early COVID-19 pandemic, the British Society of Gastroenterology (BSG) developed a risk stratification grid to inform the United Kingdom (UK) government regarding strict social isolation, termed “shielding”. This advised inflammatory bowel disease (IBD) patients thought to be most clinically vulnerable to SARS-CoV-2 infection or severe COVID-19 outcomes, to stay at home and minimize face to face contact, even with household members. Those considered at highest risk included recent commencement of combination biologic and immunomodulator therapy, prednisolone ≥20mg/day, presence of comorbidities, age ≥70 years or clinically active disease in those receiving immunosuppression. Mesalazine was not considered to increase risk. An acknowledged limitation was an absence of COVID-19 risk data. This study sought to identify patient or IBD medication-related factors associated with severe outcomes from COVID-19.
Methods
PREPARE-IBD was a multi-centre observational United Kingdom (UK) cohort study including adult IBD patients (≥18 years) diagnosed with COVID-19 by PCR between 1st March 2020 and 31st August 2020. The primary outcome was severe COVID-19 defined as requirement for intensive care admission, invasive ventilation or death. We tested associations of severe outcomes with medications and other covariates using multiple logistic regression.
Results
211 patients were included from 60 UK centres. 56 of 211 patients (26.5%) met the primary outcome. Severe COVID-19 was more common in ulcerative colitis relative to Crohn’s disease patients (33.9% [37/109] vs. 18.6% [16/86], p=0.018). Shortness of breath, nausea and vomiting were more common with severe COVID-19 (p<0.001 and p=0.023 respectively). Multivariable analysis identified co-morbidities and age as associated with severe COVID-19 outcomes; odds ratio (OR [95% CI]) 1.68 (1.23-2.35) for each co-morbidity, and an OR 1.03 (1.00-1.05) with each successive year of age. Neither clinically active IBD (OR 0.58 [0.26-1.26]), non-white ethnicity (OR 1.98 [0.92-4.28]), nor prednisolone use (OR 2.42 [0.47-11.26]) were associated with increased risk. On multivariable analysis, mesalazine was associated with severe COVID-19 outcomes (OR 2.03 [1.01-4.12]). Univariable analysis identified biologics and thiopurines as protective (OR 0.38 [0.15-0.87] and 0.32 [0.092-0.86] respectively). On multivariable analysis no association of severe COVID-19 outcomes with thiopurine or biologic exposure was seen.
Conclusion
Our data provide reassurance for the continued evidence-based use of corticosteroids, immunomodulators and biologic therapies in IBD during the ongoing COVID-19 pandemic, and is consistent with an as yet unexplained association between mesalazine use and severe COVID-19 outcomes.
Letter: online search trends suggest patient concerns around immunosuppression use in inflammatory bowel disease during COVID-19 in the United Kingdom
