Analysis of 1.1 million immune-mediated inflammatory disease patients confirms greater risk of covid death or hospitalisation.
Analysis of OpenSAFELY, one of the largest population-based datasets globally, has provided further evidence that people with immune-mediated inflammatory diseases are more likely to experience adverse covid outcomes, including hospitalisation, critical care admission and death, than the general population.
This held true even after adjusting for confounders including age, sex, deprivation and smoking status and mediators such as body mass index, cardiovascular disease, diabetes and current glucocorticoid use. The research was published in The Lancet Rheumatology.
The study also found that patients taking rituximab were at greater risk of death and critical care admission than those on other targeted therapies or conventional therapies, backing up findings from previous studies. And analysis showed that people of non-white ethnicity were at increased risk compared to white people.
The OpenSAFELY dataset comprises electronic health records of 24 million people in England, of whom 17 million adults were included in the analysis. Around 1.1 million had immune-mediated inflammatory diseases, which for the analysis were categorised broadly as joint, bowel or skin inflammatory diseases.
The most common diagnosis was psoriasis, accounting for around 60% of the cohort, with rheumatoid arthritis (15.8%), psoriatic arthritis (4.7%) and ankylosing spondylitis (3.0%) the most common inflammatory joint diseases.
After adjusting for all confounders and mediators, people with immune-mediated inflammatory diseases had a greater risk of covid-related death than the general population (HR 1.15, 95% CI 1.11-1.18).
When broken down by type of inflammatory disease, the association was greatest for inflammatory joint disease (HR 1.47, 95% CI 1.40-1.54), with smaller effect sizes for skin (HR 1.12, 1.08-1.17) and bowel (HR 1.12, 1.04-1.21) disease, when compared with the general population.
Covid-related critical care admissions or death and covid-related hospital admissions were also significantly greater among those with immune-related inflammatory diseases compared with the general population.
Analysis comparing adverse covid outcomes according to type of medication taken included around 200,000 adults (which was around 17% of those with immune-mediated inflammatory diseases).
Prescribed medications were described as standard systemic therapy (including leflunomide, methotrexate, mycophenolate mofetil, ciclosporin, sulfasalazine and azathioprine) or targeted therapies (TNF inhibitors, IL-17 inhibitors, IL-12/IL-23 inhibitors, IL-6 inhibitors, B-cell depletion therapy and JAK inhibitors).
There was no difference in adverse covid outcomes between most targeted therapies and standard systemic therapies. However, people taking rituximab had one-and-a half to two times greater risk of covid-related death, critical care admission or death, and hospitalisation than those on conventional therapies, although this was attenuated by excluding people with haematological cancer and organ transplants.
People taking JAK inhibitors had an increased risk of hospital admission compared with those on conventional therapies. The authors suggested that an unmeasured confounder explaining associations between rituximab and JAK inhibitors and covid-related adverse events couldn’t be ruled out.
In both the general population and the immune-mediated inflammatory disease cohorts, people of non-white ethnicity, such as South Asian and African descent, were more likely to experience adverse covid outcomes, with hazard ratios greater than 1.5 and more than double in some analyses compared with white people.
The key strengths of the study were the large numbers and completeness of data. Limitations include potential unmeasured confounders, in particular disease severity and shielding (for example, staying at home).
The cohort was followed from March 2020 to September 2020, so was therefore pre-vaccination, and the authors pointed out that vaccines and current treatments may mitigate some of the risks they described.
However, they suggested the data can be used to inform policy on booster vaccine prioritisation and “support health care professionals engaging in shared decision making and communication of risk”.
In an accompanying comment, Associate Professor Philip Robinson and Dr Jinoos Yazdany wrote: “We can conclude from this and other studies that patients with immune-mediated inflammatory diseases are at an increased risk of poor outcomes definitely attributed to their comorbidities and also potentially to a degree their underlying disease and specific immunosuppressive drugs.”
In light of the findings on race and ethnicity, Professor Robinson and Dr Yazdany outlined measures rheumatologists and other health professionals can take to help address covid health disparities linked with disadvantage and poor health literacy.
Actions included counselling patients to have covid tests on hand, preventative measures, pre-exposure prophylaxis for patients at high risk of poor outcomes, such as those taking rituximab, and covid treatment options.
