Pandemic reopening reshaped pediatric respiratory pathogen positivity by age

COVID-19 control measures were followed by sharply different recoveries in respiratory pathogen positivity among children treated at two Guangzhou hospitals, with influenza A, respiratory syncytial virus, influenza B and Mycoplasma pneumoniae each displaying distinct age-related patterns. A retrospective period-comparison study analyzed 230,333 patient-period records representing 197,630 children tested from January 2018 through December 2023. After strict pandemic controls, influenza A and M. pneumoniae rose above pre-pandemic positivity, while RSV increased chiefly among younger children and influenza B rebounded mainly in adolescents. The results describe test positivity among hospital attendees, not community incidence, and the researchers emphasized that changing testing practices and clinical case mix complicate direct comparisons.

The investigators divided the six-year observation window into four periods: a pre-pandemic baseline from January 2018 through January 2020; Pandemic-I, from February 2020 through June 2021, when Guangzhou used stringent measures including lockdowns, school closures, masking and travel restrictions; Pandemic-II, from July 2021 through November 2022, when restrictions were relaxed and intermittent Omicron outbreaks occurred; and a post-pandemic period from December 2022 through December 2023, after full reopening. They tested associations using multivariable logistic regression, comparing positivity in each later period with the pre-pandemic baseline while adjusting for age, sex, inpatient or outpatient status and hospital campus. Age-period interactions were evaluated with robust Wald tests, and analyses were repeated in some form to assess the influence of repeated observations and changing denominators.

Most respiratory viruses were detected less often during Pandemic-I, but the suppression was neither universal nor equally durable. Influenza A positivity fell from 19.2% before the pandemic to 5.0% during Pandemic-I, and influenza B declined from 6.5% to 0.5%. RSV fell from 21.8% to 12.3%, adenovirus from 18.4% to 1.1%, and M. pneumoniae from 12.4% to 2.2%. Rhinovirus, cytomegalovirus and bocavirus were notable exceptions to the broader viral decline. When controls were relaxed, influenza A positivity rose to 25.0%, already above its baseline, and reached 29.7% after reopening. M. pneumoniae increased more slowly, reaching 7.4% during Pandemic-II before surging to 24.1% in the post-pandemic period, the highest rate recorded for that pathogen across the four periods.

Adjustment for the changing clinical population sharpened the principal contrasts. Relative to pre-pandemic levels, influenza A had an adjusted odds ratio of 0.23 during Pandemic-I, 1.26 during Pandemic-II and 1.52 after reopening. M. pneumoniae remained below baseline during Pandemic-II, with an adjusted odds ratio of 0.54, but reached 1.99 after reopening despite a 9.3-fold expansion in the number tested. RSV produced a particularly instructive result: its crude post-pandemic positivity rate of 12.8% remained below the 21.8% baseline, yet its adjusted odds ratio was 1.31, with a 95% confidence interval of 1.06 to 1.63. Comparable-denominator analyses retained the post-pandemic increases for RSV, influenza A and M. pneumoniae, indicating that expanding testing alone did not fully account for the changes.

Age-stratified models showed that these pathogen-level averages concealed substantial differences between younger and older children. RSV positivity after reopening was elevated among children aged 29 days to 23 months, with an adjusted odds ratio of 1.50, and among those aged 2 to 4 years, where the estimate was 1.86. No significant post-pandemic increase was detected in school-aged children or adolescents, while some estimates could not be calculated because of sparse events. Influenza A increased after reopening in every age group with estimable results except neonates aged 0 to 28 days. Its largest post-pandemic estimate occurred among children aged 5 to 9 years, at 1.74, followed by 1.42 among adolescents aged 10 to 17 years.

Influenza B followed an unusual split between younger and older children. Positivity remained below baseline after reopening among children aged 29 days to 23 months and 2 to 4 years, with adjusted odds ratios of 0.53 and 0.58, respectively. In contrast, it increased among children aged 5 to 9 years, with an adjusted odds ratio of 1.21, and rose more strongly among adolescents aged 10 to 17 years, reaching 2.88. The difference between the overall and age-specific results was significant for all four principal pathogens, with age-period interaction P values below 0.001. M. pneumoniae showed a delayed increase concentrated in school-aged children: its post-pandemic adjusted odds ratio was 3.32 at ages 5 to 9 and 1.74 at ages 2 to 4, with no significant increase among infants or adolescents.

Outpatient and inpatient sensitivity analyses indicated that care setting also shaped the apparent recovery in some pathogens. Influenza A was elevated overall during Pandemic-II, but its adjusted odds ratio in inpatients was 0.64, suggesting that the early rebound was driven primarily by outpatient testing. Influenza B remained suppressed among inpatients, including an adjusted odds ratio of 0.15 after reopening, while its strong post-pandemic increase among adolescents suggested that the adolescent resurgence was concentrated in outpatient care. Because inpatient testing was used as a proxy for more consistent clinical indications, and because testing indications were not systematically recorded, the authors cautioned that residual confounding could remain even after adjustment and restriction.

Common bacterial culture results followed a different and more uncertain pattern from the viral findings. Staphylococcus aureus positivity increased from 2.7% before the pandemic to 4.2% during Pandemic-I and remained at 3.6% after reopening; its adjusted odds ratios were elevated in all three later periods. Escherichia coli also had elevated adjusted odds across the later periods, while Streptococcus pneumoniae declined after reopening, with an adjusted odds ratio of 0.77, and Pseudomonas aeruginosa reached 0.61. These estimates remain difficult to interpret because the composition of cultures changed markedly. Upper-respiratory specimens rose from 91.9% before the pandemic to 99.2% after reopening, while lower-respiratory specimens had much higher culture yields. Hospitalization rates, prior antibiotic exposure, specimen quality and reasons for obtaining cultures also varied, and some bacterial effect estimates reversed when specimen type was included in the model.

Co-infection patterns shifted as the dominant pathogens changed. During Pandemic-I, bacterial-bacterial combinations represented 49.0% of co-infections in the study’s main analysis, a period when viral positivity was at its lowest. In a sensitivity analysis excluding dual influenza A and B antigen results, bacterial-bacterial combinations declined from 62.0% during Pandemic-I to 11.2% after reopening, while bacterial-influenza combinations increased substantially and reached 22.9% after reopening. Influenza A-B dual antigen positivity accounted for 49.7% of all co-infections overall and was particularly common among children aged 5 to 9 years and adolescents, but the investigators warned that cross-reactivity in rapid antigen testing could inflate this category. Cytomegalovirus and Epstein-Barr virus detections also required caution because they may represent reactivation rather than acute respiratory infection.

The study supports an interpretation of disruption followed by non-uniform recovery, not a simple return to the old epidemiological baseline. It also cannot establish the cause of the changes. An immune gap, altered transmission, shifts in social mixing and changing patterns of healthcare use are plausible explanations discussed by the authors, but no serological or immunological measurements were collected, so the immune-gap mechanism remains a hypothesis. Other limitations include the retrospective design, restriction to two referral hospitals in one city, changing test denominators, lower sensitivity of rapid antigen assays for influenza, sparse data in some age-specific strata and differences in period lengths that prevented a strict interrupted time-series analysis. The investigators concluded that the divergent results support consideration of pathogen- and age-specific surveillance, while describing the findings as observations among tested children rather than direct measures of population incidence or evidence that any prevention strategy would prevent infection.

Subject of Research: Age-specific changes in pediatric respiratory pathogen test positivity across COVID-19 pandemic periods in Guangzhou

Article Title: Disruption and change in positivity of respiratory pathogen epidemiology in children during and after the COVID-19 pandemic, 2018–2023: a large-scale period-comparison study in Guangzhou, southern China

Article References: Li, H., Bai, R., Tang, Z., Li, P., Tang, L., Zheng, X., Gu, C., Ji, C., Luo, Y., & Luo, M. (2026). Disruption and change in positivity of respiratory pathogen epidemiology in children during and after the COVID-19 pandemic, 2018–2023: a large-scale period-comparison study in Guangzhou, southern China. Frontiers in Microbiology, 17, Article 1992791. https://doi.org/10.3389/fmicb.2026.1992791

Image Credits: AI Generated

DOI: 10.3389/fmicb.2026.1992791

Keywords: COVID-19 pandemic, pediatric respiratory infections, influenza A, influenza B, respiratory syncytial virus, Mycoplasma pneumoniae, pathogen positivity, age stratification, Guangzhou, post-pandemic epidemiology, Disruption, change