Evaluation of a rapid antigen detection test in the diagnosis of streptococcal pharyngitis in children and its impact on antibiotic prescription

To study the performance of the Becton-Dickinson Link 2 Strep A Rapid Test, a rapid antigen detection test (RADT) for diagnosing streptococcal pharyngitis in children presenting to private offices and to the Pediatric Outpatient Clinic of a university hospital, in relation to clinical criteria (fever, tender anterior cervical lymph nodes, tonsillar exudate and absence of cough), and its impact on antibiotic prescription. Children were enrolled in Group A (enrolment by private-practice paediatricians; diagnosis by clinical picture only), Group B (enrolment by private-practice paediatricians; diagnosis by RADT and culture) or Group C (enrolment by hospital-affiliated paediatricians in the Pediatric Outpatient Clinic; diagnosis by RADT and culture). During a 2 year period, 820 children were enrolled [369 (45%) in Group A, 270 (33%) in Group B and 181 (22%) in Group C]. Streptococcal pharyngitis was diagnosed by RADT and culture in 146 (32.4%) of the 451 tested children. The sensitivity, specificity and positive and negative predictive values of the RADT were 83.1%, 93.3%, 82.4% and 93.6%, respectively. A stepwise increase in the sensitivity of the RADT was noted among children with one, two, three or four clinical criteria (60.9% to 95.8%). Paediatricians without access to laboratory tests were more likely to prescribe antibiotics compared with paediatricians with access to tests (72.2% versus 28.2%, P < 0.001). Private-practice paediatricians prescribed antibiotics more frequently compared with hospital-affiliated paediatricians (55.7% versus 19.9%, P < 0.001). Our findings support screening of all children with pharyngitis for Centor criteria and subsequently performing an RADT to guide decision for antibiotic administration. Such a strategy has an important impact on limiting throat culture testing and is associated with reduced antibiotic prescription.