Faculty of Public Health - Andalas University - OCS, 13th IEA SEA Meeting and ICPH - SDev

Font Size: 
Validation of clinical scores for identification of bacterial sore throat among pediatric outpatient attendees in North India
Raghavan Parthasarathy

Last modified: 2018-09-16


Background: Streptococcal sore throat leads to sequelae of Rheumatic Heart Disease (RHD) in children. One of the ways to prevent RHDs is primary prophylaxis by treating all bacterial sore throats by antibiotics, which raises the issue of avoidable antibiotic use. A clinical score would be an inexpensive and cost-effective alternative to bacterial culture in guiding bacterial sore throat management and in preventing RHD.

Materials and methods: The study was among pediatric (5-15 years old) outpatient attendees of a sub-district hospital, in North India in 2017. Children reporting with sore throats were identified and enrolled in the study and a full clinical examination was conducted and throat swab taken. Five clinical scores (Centor, McIsaac, Bakare, Breese and a new proposed score) were validated against bacterial culture. Receiver Operating Characteristic (ROC) curves were plotted for various clinical scores and Area Under the Curve (AUC) estimated. Sensitivity, specificity with 95%CI were estimated for most appropriate cut-offs of the score.

Results: The prevalence of bacterial sore throat among sore throat cases was 4.30% (95%CI: 2.94-6.05). Sensitivity was highest for Bakare (90.3%; 95%CI: 87.9-92.3%), and least for new score (74.2%; 95%CI: 70.8-77.3%). Specificity was highest for McIsaac, 51.8% (95%CI: 48.1-55.5%), and least for Bakare (38.8%; 95%CI: 35.3-42.6%). AUC was best for the Bakare score (73%; 95%CI: 68-75%).

Conclusion: Available clinical scores are useful and have high sensitivity but only modest specificity. Their application could help in rationalization of antibiotic usage in the primary care settings and prevent streptococcal sequelae.

Keywords: Sore throat, clinical score, bacterial, validation, antibiotics