Temporal Trends of Serum Epstein-Barr Virus DNA Positivity Among Clinically Tested Pediatric Patients in Hangzhou, China, 2016-2025: A 10-Year Retrospective Study.

This study aimed To describe the 10-year temporal trends of serum Epstein-Barr virus (EBV) DNA positivity among pediatric patients tested at Hangzhou Children's Hospital, and to explore demographic and seasonal variations across pre-pandemic, pandemic, and post-pandemic periods. A total of 11,689 serum samples from children aged 0-15 years who underwent EBV nucleic acid testing at Hangzhou Children's Hospital from January 2016 to December 2025 were retrospectively analyzed, including 2,160 positive results. Demographic characteristics (sex, age), visit type (outpatient/inpatient), and temporal distributions (year, month) were examined. Monthly detection rates were calculated using monthly testing denominators. Group comparisons were performed using chi-square tests; the linear-by-linear association Chi-square test was applied to annual rates. Multivariable negative-binomial regression with offset for monthly test volumes and interrupted time-series analysis (ITSA) were used to evaluate independent associations with secular trend, seasonality, and pandemic phase. The overall EBV-DNA detection rate was 18.48% (2,160/11,689) among tested samples. A deduplication sensitivity analysis yielded a corrected patient-level rate of 18.26%, confirming robustness. Females (20.04%) had a higher rate than males (17.26%, χ2 = 14.66, p < 0.01). Outpatients (28.24%) showed a higher rate than inpatients (16.62%, χ2 = 141.6, p < 0.01). Age distribution was concentrated in children aged 1-8 years (86.02% of positives), peaking at 5 years (25.14%) and lowest in infants < 1 year (2.46%). Annual detection rates rose from 7.25% (2016) to 29.69% (2025) (linear-by-linear association Chi-square test, p < 0.01). Monthly detection rates suggested a descriptive bimodal pattern, with an autumn peak (September-November, 21.07%) and a spring peak (March-May, 17.88%), although formal testing did not reach statistical significance (joint Wald χ2 = 14.72, df = 11, p = 0.195). Across pandemic phases, unadjusted rates were 12.03% (pre-pandemic), 19.61% (pandemic), and 21.64% (post-pandemic). Multivariable negative-binomial regression revealed that the long-term secular trend was the only significant predictor (adjusted rate ratio [aRR] = 1.029 per month, 95% CI: 1.000-1.059, p = 0.047), while pandemic phase (aRR = 1.171, p = 0.347) and post-pandemic phase (aRR = 1.233, p = 0.288) were not significant. ITSA showed no significant level or slope changes at either breakpoint (all p > 0.05). In this hospital-based tested population, serum EBV-DNA positivity showed a significant upward secular trend over 2016-2025. The post-pandemic increase was primarily driven by a pre-existing secular trend rather than a discrete "rebound" attributable to immunity debt. Due to selection bias from non-standardized testing indications, these findings should not be generalized to the general pediatric community. Future studies with population-based sampling and serological differentiation of primary infection versus reactivation are needed.