Use and predictive value of the determination of antinuclear antibodies in a national health reference hospital

A retrospective review of the clinical and demographic data was made in a teaching hospital of 950 beds. A consecutive sample of inpatients and outpatients with medical request of indirect immunofluorescence antinuclear antibodies test (IIF-ANA) in a period of 18 months was taken to define the positive predictive value (PPV) and the negative predictive value (NPP) of the IIF-ANA, anti-ds DNA antibodies, and anti-extractable nuclear antigens (anti-ENA) for the diagnosis of SLE and connectivopathies. The PPV and the NPP of the IIF-ANA for the detection of anti-DNA antibodies and anti-ENA were also calculated. It was found that 2 113 patients had been referred for the determination of IIF-ANA, of whom 273 were determined anti-doble strand DNA antibodies (anti-ds DNA) and 233 anti-extractable nuclear antigens (anti-ENA). 651 were registered with diagnosis of connective tissue disease (CTD), and 97 of them with systemic lupus erythematosus (SLE). The predictive values of the positive results of IFI-ANA for the CTD and SLE were 59.8 and 22.8 %, respectively, whereas this same value of the anti-ENA for the CTD was 99.0 %, and that of the anti-ds DNA for the SLE was 97.3 %. The predictive values of the negative results of the IFI-ANA for the CTD and the SLE were 76.0 and 99.7 %, respectively. The predictive value of the negative results of the anti-ENA for the CTD was 28.4 % and that of the anti-ds DNA for the SLE was 87.9 %. It was concluded that most of the results of the IIF-ANA were negative and they were required for patients without connectivopathies, which gave rise to low predictive values and to a questionable clinical usefulness. These data suggest an inadequate overuse of the IIF-ANA test out of a logical clinical context. This situation may be corrected with a careful pre selection of the patients based on the presence of various recognized clinical criteria of the connectivopathies