Blood And Pharyngeal Swab Polymerase Chain Reaction Testing: Diagnostic and Therapeutic Implications in Children With Acute Nonbacterial Encephalitis

Background: Children with acute nonbacterial encephalitis are often treated with extended antimicrobial therapies. Negative results from cerebrospinal fluid (CSF) examinations frequently lead to unnecessary prolongation of empirical treatments due to uncertain etiology. To enhance the identification rate of infectious causative agents, additional tests are increasingly performed on non-CSF specimens, including viral polymerase chain reaction (PCR) testing on blood or pharyngeal swabs. This study aimed to evaluate the impact of blood and pharyngeal swab PCR testing on the duration of empirical intravenous (IV) therapy in children with acute nonbacterial encephalitis, as well as its usefulness in identifying viral etiologies. Methods: A single-center, retrospective observational study was conducted on children aged 28 days to 18 years hospitalized with acute, nonbacterial encephalitis from January 2019 to August 2024. Data on age, sex, clinical presentation, laboratory features, instrumental investigations, administered therapy, results of PCR testing on different specimens and identified etiology were collected. Results: Among 52 encephalitis cases, viruses were detected via PCR in CSF in 9 cases (17%), pharyngeal swabs in 8 (15%) and blood in 5 (10%). Herpetic infections were more frequently found on CSF (P < 0.001), while nonherpetic viruses were detected in non-CSF specimens in most cases. The duration of IV antibiotics was shorter in patients with positive viral PCR results from blood or pharyngeal swabs (P = 0.009). Prolonged antibiotic therapy despite negative bacterial CSF tests was inversely associated with positive PCR in blood/pharyngeal swabs [adjusted odds ratio (aOR), 0.14; P = 0.031) and directly associated with sepsis-like presentation (aOR, 5.08; P = 0.028). No difference in IV acyclovir duration was found between those with nonherpetic viral infections and those with unidentified etiology (P = 0.12). Conclusions: Performing viral PCR on both CSF and non-CSF specimens is associated with a higher detection rate of viral etiology and may positively impact the duration of empirical antimicrobial therapy in children with nonbacterial encephalitis.