Introduction/Aims: Antinuclear antibody (ANA) tests have good sensitivity but poor specificity for autoimmune rheumatic diseases. ANAs are frequently misused as a screening test without considering the pre-test probability of ANA-related disease (ARD). This leads to false positives that confuse clinical decisions and trigger unnecessary and expensive downstream ANA-related tests. Previous studies have reported the low predictive value of ANAs mainly in outpatient settings. However, ANAs performed on hospital inpatients have been poorly described so far. We examined the clinical indications and positive predictive value (PPV) of inpatient ANA testing. We also quantified the costs of ANA and ANA-related tests.
Methods: We retrospectively analysed 105 inpatients over 18 years old who had a positive ANA result at a major tertiary hospital between January 1st to February 28th, 2023. The reason for ordering the ANA, ordering subspecialty, and any subsequent ARD diagnosis were recorded. ANA-related tests such as ENA and dsDNA performed in the subsequent six months after the initial positive ANA was recorded. The cost of ANA and ANA-related testing was analysed.
Results: There were 117 positive ANA tests with only 2 ARDs diagnosed, resulting in a low PPV of 1.7%. The top reasons for testing were organ-related pathologies such as Hepatic (n=38), Neurological (n=16), and Haematological (n=13) issues. Among the various specific reasons under 'Hepatic', most were reflexive screens for incidental undifferentiated LFT derangement (n=27). The tests consumed $14008.55, indicating a cost of $7004.28 per diagnosis, given a positive ANA.
Conclusions: This is the first study examining the clinical utility of ANAs in an inpatient Australian setting, and the first study to quantify the costs of ANA-related tests. Our main new finding is that the reasons for ordering ANAs in hospital inpatients are mainly organ-related, in contrast to the musculoskeletal reasons in outpatient settings reported in previous studies. This could explain our low inpatient PPV, because ARDs are chronic conditions that likely first present with outpatient musculoskeletal issues instead of inpatient organ issues. This paper also reveals the extremely high cost of inpatient ANA and related tests, likely higher if we also accounted for negative ANAs. As hepatic reasons were the most common testing indication, for future research we suggest the possible development of an ‘ANA Hepatic Score'. It can use the risk factors for autoimmune hepatitis to assess pre-test probability before ANAs are ordered, somewhat similar to the purpose of the Wells Score. This could materially reduce inpatient ANAs.