Group A Streptococcal Tonsillopharyngitis: The Diagnostic Power of the Centor and McIsaac Clinical Prediction Models at Different Pre-probability

Abstract:

Background: The power of diagnostic tests is affected by pre-test probability, and clinical prediction models must be validated in different populations. The aim ofthis study was to determine the diagnostic value of symptoms and signs for group A Streptococcus tonsillopharyngitis and diagnostic power of Centor and Mclsaac crite- ria in a patient population with different pre-probability.

Methods: The study was conducted between September 2019 and February 2020 in Adnan Menderes University Hospital’s outpatient clinics. A total of 405 patients olderthan 36 months who presented with one of the complaints of acute tonsillopharyn-gitis participated in the study. Throat swab samples were taken from each patient. The diagnostic value of symptoms and signs was determined by performing univariateanalysis and multiple logistic regression analysis.

Results: The mean age of 405 patients was 24.7 (3-81 years). While group AStreptococcus positivity was 7.9% over the age of 3, the frequency of group A Streptococcus was 16.8% in children under the age of 15 and 4.7% in adolescents and adults. Group A Streptococcus positivity was 45.8% in those with a Centor score of 4 and 35.7% in those with a Mclsaac score of 4-5. In regression analysis, only 4 criteria included in the Centor score entered the model (P < .05).

Conclusions: Centor and McIsaac clinical prediction models were found to be valid inour patient group with low group A Streptococcus positivity. However, although thediagnostic power of both clinical prediction models does not change in the patient population with low group A Streptococcus positivity, they cannot increase the post- test probability above 40-50%.