P015 Improving the diagnostic performance of ANCA testing: evaluating the use of a gating strategy for ANCA test requests

Abstract Background/Aims  Anti-neutrophil cytoplasmic antibody (ANCAs) associated vasculitis (AAV) encompasses: Granulomatosis with polyangiitis (GPA), microscopic polyangiitis (MPA) and eosinophilic granulomatosis with polyangiitis (EGPA). The 1999 international consensus for ANCA testing recommended the use of indirect immunofluorescence (IIF) to screen for ANCAs, with immunoassays for proteinase-3 (PR3)-ANCAs and myeloperoxidase (MPO)-ANCAs in positive samples. In 2017 this consensus was revised, recommending the implementation of a strict gating strategy for ANCA requests, followed by high-quality immunoassays as the primary screening method, to reduce the volume of tests requested, reducing the false-positive rates, thereby improving the diagnostic performance of ANCA testing. Testing at the Royal Wolverhampton Trust is based on the 1999 consensus. Tests perform poorly when not carried out in the correct setting due to low pre-test probability; we aimed to implement a gating strategy to filter inappropriate ANCA test requests to improve the diagnostic performance of ANCA testing at our centre. We assessed the use of IIF and antigen specific assays in the diagnosis of AAV. Methods  We reviewed the records of patients that had ANCA tests requested in June 2019. A gating strategy in the form of an indication selection ‘pop-up box’ directing clinicians to select an appropriate clinical indication for the request was implemented. We re-audited ANCA requests made post-intervention Feburary 2020. ANCA requests were assessed against clinical indications in the 1999 consensus for appropriateness; we assessed the diagnostic performance by the proportion of ANCA tests correctly identifying AAV and calculating the sensitivity and specificity of IIF and immunoassays for MPO/PR3-ANCAs in detecting AAV. Results  Post-intervention, 105 fewer ANCA requests were made (298v403). Due to insufficient information, 345 patients pre-gating and 252 patients post-gating-strategy were used for analysis. The proportion of appropriate requests made was greater post-intervention, 215/252 (85.3%) vs 119/345(34.5%). Pre-intervention AAV was identified in 13 (3.8%) patients; AAV was identified in 21 (8.3%) patients following the gating-strategy introduction. IIF was found to have sensitivity of 91.7% pre and 76.2% post-intervention; specificity was 53.0% pre and 58.9% post-intervention. MPO/PR3 immunoassay testing was found to have a sensitivity of 91.7% pre and 57.1% post-intervention; specificity was 97.4% pre and 97.9% post-intervention. Conclusion  The introduction of a gating-strategy for ANCA testing at our centre effectively reduced the overall number of ANCA requests by 26.1% over a 1-month period and significantly increased requests with an appropriate recorded indication. The proportion of tests correctly identifying AAV diagnosis increased following the introduction of the gating strategy indicating an improvement in diagnostic performance of ANCA testing. These results indicate antigen specific testing is more specific for a diagnosis of AAV and sensitivity of IIF was reduced following introduction of the gating strategy suggesting IIF is of no additional benefit when an effective gating strategy is employed. Disclosure  N. Cleaton: None. N. Barkham: None. T. Adizie: None.