Treatment outcomes of acute streptococcal tonsillitis according to antibiotic treatment. A retrospective analysis of 242,366 cases treated in the community

BACKGROUND: Acute upper respiratory infections are the most common reason for primary physician visits in the community. This study investigated whether the type of antibiotic used to treat streptococcal tonsillitis can reduce the burden by affecting the number of additional visits. OBJECTIVES: To assess the effect of different antibiotic treatments for tonsillitis on the number of additional primary physician visits and the development of infectious or inflammatory sequels. METHODS: This retrospective study included first cases of culture-confirmed streptococcal tonsillitis (n = 242,366, 55.3% females, 57.6% aged 3–15 years) treated in primary clinics throughout Israel between the years 2010 and 2019. Primary outcomes were the number of additional primary physician visits, due to any cause or due to specific upper airway infections. Secondary outcomes were the number of developed complications, such as peritonsillar abscess, post-streptococcal glomerulonephritis, rheumatic fever, post-streptococcal arthritis, chorea and death. RESULTS: Compared to penicillin-V, adjusted incidence rate ratios (IRR) for additional primary physician visits at 30–days were highest for IM benzathine-benzylpenicillin (IRR = 1.46, CI 1.33–1.60, p < .001) and cephalosporin treatment (IRR = 1.27, CI 1.24–1.30, p < .001). Similar results were noted for visits due to specific diagnoses such as recurrent tonsillitis, otitis media and unspecified upper respiratory tract infection. Amoxicillin showed decreased adjusted odds ratio (aOR) of developing complications (aOR = 0.68, CI 0.52–0.89, p < .01 for any complication. aOR = 0.75, CI 0.55–1.02, p = .07 for peritonsillar or retropharyngeal abscess). CONCLUSION: Penicillin-V treatment is associated with fewer additional primary physician visits compared to other antibiotic treatments. Amoxicillin and penicillin-V are associated with fewer complications. These findings are limited by the retrospective nature of the study and lack of adjustment for illness severity. Further prospective studies may be warranted to validate results.