Objective: To assess the effect of a simple intervention on antinuclear antibody (ANA) test overuse by rheumatologists.
Methods: This was an explorative, pragmatic, before-and-after, controlled implementation study among rheumatologists working at 3 rheumatology departments in secondary and tertiary care centers in The Netherlands. The intervention was given in all study centers separately and combined education with feedback. Six outcome measures describe the intervention effects: the ANA/new patient ratio (APR), difference with the target APR, percentage of positive ANA tests, percentage of repeated ANA testing, percentage of ANA-associated diseases, and APR variation between rheumatologists. Outcomes were compared between the pre- and postintervention period (both 12 months) using (multilevel) logistic regression or F testing. Results are reported together for centers 1 and 2, and separately for center 3, because ANA tests could not be linked to an individual rheumatologist in center 3.
Results: The APR decreased from 0.37 to 0.11 after the intervention in centers 1 and 2 (odds ratio [OR] 0.19, 95% confidence interval [95% CI] 0.17-0.22, P < 0.001) and from 0.45 to 0.30 in center 3 (OR 0.53, 95% CI 0.45-0.62, P < 0.001). The percentage of repeated ANA requests in all centers and the APR variation for centers 1 and 2 decreased significantly. Only in center 3 did the percentage of ANA-associated diseases increase significantly.
Conclusion: A simple intervention resulted in a relevant and significant decrease in the numbers of ANA tests requested by rheumatologists, together with an improvement on 3 other outcome measures.
Trial registration: ClinicalTrials.gov NCT02409251.
© 2016, American College of Rheumatology.