Optimizing diagnosis of pediatric Streptococcus pyogenes infections: A clinical prediction model

Group A Streptococcus (GAS) infections, caused by Streptococcus pyogenes , are a major health concern among children under 12 years, leading to conditions such as tonsillitis, rheumatic fever, and post-streptococcal glomerulonephritis. Accurate diagnosis is essential for effective treatment and prevention. This study aimed to develop and validate a Clinical Prediction Rule (CPR) for predicting the presence of S. pyogenes in throat swabs from children under 12. We analyzed clinical and laboratory data from 1015 pediatric patients presenting with symptoms suggestive of GAS infection at Federal University Teaching Hospital, Owerri, Imo State, between January 2019 and December 2022. S. pyogenes was first isolated from throat swabs cultured on blood agar, and 233 positive isolates were further identified using Polymerase Chain Reaction (PCR) targeting the Spy 1258 gene, with DNA extracted via boiling and confirmed through phenotypic methods. Data from these 233 identified cases were used to develop and validate the CPR. Variables examined included gender, age, ward of admission, clinical diagnosis, and antibiotic susceptibility. Logistic regression modeling identified significant predictors of S. pyogenes presence, with potential biases minimized through systematic case review, standardized data extraction and cross-checking by multiple reviewers. Among the 233 cases analyzed, the mean age was 4 ± 0.25 years, with 62.7 % under age 3. Tonsillitis was the predominant diagnosis, with GAS prevalence ranging from 56.7 % to 69.3 %. Antibiotic susceptibility test result varied, with significant predictors including Sepsis/Tonsillitis and ear discharge/tonsillitis (p < 0.001). The CPR model demonstrated a sensitivity of 95.4 % and a specificity of 36.8 %, highlighting its potential to enhance clinical diagnosis and management of GAS infections. This study offers valuable insights into predictors of S. pyogenes infection in pediatric patients and highlights the CPR’s potential for improving clinical diagnosis and management. • Development of Clinical Prediction Model: Created and validated a Clinical Prediction Rule (CPR) for predicting S. pyogenes infections in children under 12 years using clinical and laboratory data. • High Sensitivity and Moderate Specificity: The CPR model showed high sensitivity (95.4 %) for S. pyogenes infections but moderate specificity (36.8 %) for non- S. pyogenes infections. • Significant Predictors: Sepsis/Tonsillitis and Ear Discharge/Tonsillitis are significant predictors of S. pyogenes infection (p < 0.001). • Antibiotic Susceptibility Insights: variable effectiveness found, with Erythromycin and Chloramphenicol effective, but resistant to Amoxicillin and Cefotaxime. • Implications for Clinical Practice: The CPR model is a valuable tool for enhancing diagnosis and management of S. pyogenes infections, improving patient outcomes and guiding antibiotic stewardship in pediatric.