Antinuclear antibody (ANA) testing. Why did I request it and what do I do now? An observational study of UK primary care.

Conference:

SAPC ASM 2023 - Brighton

Talk Code:

7E.4

Presenter:

Dr Will Evans

Twitter:

@willrhevans

Co-authors:

Will Evans, Yana Vinogradova, Fiona Pearce, Peter Lanyon

Author institutions:

NIHR School of Primary Care Research School of Medicine University of Nottingham; Division of Epidemiology and Public Health, University of Nottingham

Problem

The ANA-associated diseases are a group of rare autoimmune diseases (including: SLE, Sjögren’s, Scleroderma, autoimmune hepatitis & myositis). These diseases are hard to diagnose, and patients often experience a long diagnostic delay.The ANA test, frequently performed in primary care can be difficult to interpret, it is often positive in patients who do not appear to have an associated disease at the time of testing, although a proportion will later develop disease. This presents a challenge: How should these patients be followed-up? We suspect there may be features in these ANA positive patients that can help distinguish those who will later develop disease.To address these challenges, we aim to:• firstly, understand the characteristics of those who are tested;• secondly, compare the incidence of ANA-associated disease and mortality between the tested and untested populations;• and finally, in those who are positive identify features that are associated with later ANA disease development.

Approach

We have collected from the literature, disease experts and patients, features that prompt ANA testing and features associated with disease development. We have generated code lists for these features. We have used the primary care database CPRD Aurum to perform the following observational studies. We have linked hospital and mortality data to enhance the capture of outcomes. The study involves 3 phases:1) A parallel cohort study comparing the ANA-tested population subdivided by test result (positive, negative or unknown result) to the untested population. This will the enable Incidence rates for the outcomes: a composite of ANA-associated diseases and mortality, to be compared between exposure groups.2) A case control study to understand what led to ANA testing. Each patient who has had an ANA test is matched 1 to 1 to a control patient. Logistic regression will be used to calculate odds ratios for exposures associated with performing an ANA test.3) A cohort study of ANA positive patients. We will use time to event analysis with Cox- regression models to calculate hazard ratios for candidate predictor variable for later ANA disease development.

Findings

At the time of completing this abstract submission the code-lists have been generated, the data has been cleaned and almost completely prepared for analysis. The analyses have not yet been performed but will be over the next 3 months.

Consequences

We envisage that the findings of this study will help to stratify patients who test ANA-positive and help reduce unnecessary referrals and target resources more effectively. It may also give insights into when to request and when not to request an ANA test, reducing the number of clinically inappropriate requests. Insights from this study may also help to improve the diagnosis of patients affected by ANA-associated diseases, a priority of the UK Rare disease Framework.

Submitted by:

Will Evans

Funding acknowledgement:

Dr Will Evans, (GP Career progression fellowship: NIHR SPCR:C019) is funded by Health Education England (HEE) / NIHR for this research project. The views expressed in this publication are those of the authors and not necessarily those of the NIHR, NHS or the UK Department of Health and Social Care.