BACKGROUND: Sore throat is a common complaint, with causes that include viruses and group A b-hemolytic streptococcus. Untreated pharyngitis due to this latter organism can lead to serious sequelae, such as peritonsillar abscess and rheumatic fever. A quick and accurate diagnosis, therefore, is important. Despite its relatively high prevalence among patients with pharyngitis, always performing a diagnostic laboratory test to uncover group A streptococcus is both impractical and costly. Identifying clinical correlates of strep throat would be useful.
POPULATION STUDIED: Adult and pediatric outpatients presenting with sore throat were studied.
STUDY DESIGN AND VALIDITY: The authors performed a thorough MEDLINE search to identify and systematically review studies of the diagnosis of group A b-hemolytic streptococcal pharyngitis in patients presenting with sore throat. Unpublished data were not sought. Initially 917 articles were identified, of which 9 met level I evidence criteria (large blinded prospective studies using throat culture as the reference standard). Pairs of authors reviewed each study, and discrepancies were resolved by discussion. The authors then pooled data to calculate the sensitivity, specificity, positive likelihood ratio (LR+), and negative likelihood ratio (LR-) of various history and physical examination elements. The authors only included studies with 300 or more subjects. Based on this threshold, 8 studies (1182 patients) were excluded from the pooled analysis.
OUTCOMES MEASURED: Primary outcomes measured included the sensitivity, specificity, LR+, and LR- of different clinical features.
RESULTS: The presence of tonsillar exudate or, pharyngeal exudate and a history of streptococcus exposure in the previous 2 weeks were most useful in predicting current streptococcus pharyngitis (LR+ = 3.4, 2.1, and 1.9, respectively). The absence of tender anterior cervical lymph nodes, tonsillar enlargement, and tonsillar or pharyngeal exudate was most useful in ruling out strep throat (LR- = 0.60, 0.63, and 0.74, respectively).
No single history or physical examination element can effectively rule in or rule out strep throat. However, clinical prediction rules using a constellation of clinical symptoms and signs (such as presence of tonsillar or pharyngeal exudate and history of exposure to streptococcus) can be helpful in diagnostic testing and treatment decisions when patients present with a sore throat in the outpatient setting. No recommendations were made regarding which probability thresholds should be used when making treatment decisions.