Does a clinical prediction rule anticipate the diagnosis for streptococcal pharyngitis in children aged 2 to 15?

Abstract

Background: The etiology of a streptococcal pharyngitis must be documented by laboratory techniques to avoid unnecessary antimicrobial treatment, but this strategy increases cost for the patient. Available scores applied in children or adults are imperfect. Aim: To develop a clinical prediction rule to aid the diagnostic process of streptococcal pharyngitis (SP) in children in a low-resource setting. Methods: Three hundred and eighteen patients aged 2 to 15 years who were evaluated for suspected SP at the Pediatric Emergency Department and the Pediatric Ambulatory Unit of Red Salud UC-Christus entered the study. A throat culture and a rapid antigen detection test for Streptococcus pyogenes were obtained from each patient. Data were analyzed for possible clinical predictors of SP with univariate and multiple regression analyses. Results: Seventy-three cases of SP were diagnosed (23.9%). In the univariate analysis, fever was inversely associated with SP (p = 0.002). Odynophagia, palatal petechiae, and season of the year (autumn and winter) were positively associated with SP (p = 0.007, p < 0.001 and p = 0.03 respectively). In multiple regression analysis the models did not have sufficient power to predict streptococcal etiology. Conclusion: Clinical predictors, even those systematically included in clinical prediction rules, did not show sufficient predictive power to safely include or exclude SP in this setting, and thus, it is necessary to improve access to confirmatory tests.