Methods
Participants Four cohorts of participants were recruited: the first group included 24 convalescent health care workers who had suffered from COVID-19 and not required hospitalisation to determine whether disease severity makes a difference to the generation of any autoantibodies. The second group included 25 individuals who were currently on the intensive therapy unit (ITU) diagnosed with COVID-19. The third group included 32 individuals who were currently on the ITU but for a reason other than COVID-19 to determine whether acute illness per se was associated with autoantibody production. The fourth group included 35 individuals who had been admitted to ITU with COVID-19 and had been discharged and their samples were taken at 3-6 month review. This group explored persistence of any potential antibodies.
Autoantibody assays Autoantibodies were detected on serum samples using indirect immunofluorescence. A spectrum of anti-neutrophil and organ specific autoantibodies were tested as a broad screen using standard ISO15189 accredited clinical tests. The immunoglobulin isotype detected was IgG with the exception of endomysial antibodies which are IgA. The full list of assays, manufacturers and disease association are described in theSupplementary data but in short we undertook indirect immunofluorescence on commercial pre-prepared slides to detect IgG antibodies for adrenal, autoimmune encephalitis, anti-neutrophil (ANA), anti-neutrophil cytoplasmic ANCA, cardiac, epidermal, islet cell, a range of cerebellar (Purkinje cell) antibodies, smooth muscle, mitochondrial, gastric parietal cell and skeletal muscle antibodies and IgA endomysial antibodies. All samples were read by two experienced clinical laboratory scientists for agreement. Data are largely descriptive. Inter group comparison was made by Chi Square using GraphPad Prism 9.
Ethical approval
For the acute cohort of ITU patient, surplus anonymised samples from routine clinical testing were used and for the convalescent ITU patients they were consented in clinic; ethical approval for these groups was granted by the North West-Preston Research Committee (ref 20/NW/0240 IRAS Project ID: 282164). The health care worker cohort was a random subgroup of COVID antibody positive patients from the COvid-19 COnvalescent immunity study (COCO) study approved by the London - Camden & Kings Cross Research Ethics Committee reference 20/HRA/1817.