The score for allergic rhinitis (SFAR): a simple and valid assessment method in population studies

BACKGROUND: No validated assessment of allergic rhinitis (AR) is presently available that can be used in population studies in the absence of medical diagnosis and of objective measurements of allergy. To compensate for this lack, a quantitative Score For Allergic Rhinitis (SFAR) ranging between 0 and 16 has been developed by experts. METHODS: The SFAR, encompassing eight features of AR, was validated in three different ways: 1) among 269 outpatients taking the specialist's diagnosis of AR and skin prick tests (SPT) positivity as a gold standard (diagnosis validation); 2) using psychometric methods (internal validation); and 3) in a random population-based sample of 3001 individuals by telephone interview (population acceptability). RESULTS: A SFAR value > or = 7 allowed satisfactory discrimination between the outpatients with AR from those without (sensitivity = 74% [95% confidence interval CI: 0.69,0.79], specificity = 83% [0.79, 0.87], positive predictive value = 84% [0.80, 0.88], negative predictive value = 74% [0.69, 0.79] and Youden's index = 0.57, respectively). Internal consistency of the score was also high (among others, Cronbach's alpha coefficient = 0.79). On average, it took only 3 min for the individuals interviewed on the phone to complete the questionnaire, the questions of which were well understood. Among these subjects, the prevalence of AR was 21% [95% CI: 19.5%, 22.5%], which is comparable to other determinations in France. CONCLUSIONS: The newly a priori proposed Score For Allergic Rhinitis (SFAR) is easy to use and can be useful to estimate prevalence and to study causation of AR in population settings.