Adenovirus and Mycoplasma Leave Distinct Biomarker Fingerprints in Hospitalized Children

When a child arrives at a hospital with fever, cough, and labored breathing, clinicians face a familiar diagnostic puzzle. A multiplex polymerase chain reaction panel can reveal which virus or bacterium is lurking in the nasopharynx, while blood tests measure the body’s inflammatory response. Yet the two kinds of information rarely line up neatly, and neither alone tells the medical team how long the child will need a hospital bed. A new retrospective cohort study from Hangzhou Children’s Hospital in China now offers one of the more detailed pictures of how these pieces fit together, drawing on more than six thousand pediatric admissions over two and a half years.

The study, published in BMC Infectious Diseases, was led by Yalin Sun and colleagues in the hospital’s Department of Pediatric Infectious Diseases. The team examined children aged eighteen years or younger who were hospitalized with acute respiratory tract infections between 5 January 2023 and 17 June 2025. Each child underwent a thirteen-target pharyngeal-swab assay covering the common respiratory pathogens, from respiratory syncytial virus and influenza to human adenovirus and Mycoplasma pneumoniae. The researchers collapsed these detections into eleven analytical groups and then asked a deceptively simple question: did the identity of the detected pathogen predict the child’s inflammatory biomarker levels or the length of their hospital stay?

The scale of the dataset lends the findings considerable weight. Of 6,093 hospitalized children included in the analysis, 2,076 tested negative on the respiratory panel, 3,343 had a single pathogen detected, and 674 carried detections of multiple targets simultaneously. Rather than treating every pathogen individually, the investigators defined three primary outcomes that map onto everyday clinical decisions: a high-sensitivity C-reactive protein value of at least 20 milligrams per liter, a procalcitonin value of at least 0.5 nanograms per milliliter, and a hospital stay lasting seven days or longer. These thresholds are commonly used at the bedside as rough markers of a bacterial process or a complicated course.

The statistical approach was deliberately conservative. Complete-case logistic regression models adjusted for age group, sex, testing year, and season, and the authors applied false discovery rate corrections to guard against spurious associations arising from the many comparisons inherent in a multi-pathogen panel. Sensitivity analyses probed the robustness of the results from several angles: biomarkers were also modeled as continuous variables, multiple-target detections were examined separately, recorded SARS-CoV-2 positivity was accounted for, and the analysis was restricted to children sampled within 24 or 48 hours of admission using both source records and chart-reviewed timing intervals.

Two pathogens stood out with sharply contrasting signatures. Human adenovirus detection was associated with roughly double the odds of an elevated inflammatory response compared with children whose panels came back negative. The adjusted odds ratio was 2.19 for a high-sensitivity C-reactive protein value of at least 20 milligrams per liter, with a 95 percent confidence interval of 1.63 to 2.96, and 1.93 for a procalcitonin value of at least 0.5 nanograms per milliliter, with a confidence interval of 1.34 to 2.77. In practical terms, a child whose swab revealed adenovirus was markedly more likely to show the kind of inflammatory surge that clinicians often associate with bacterial infection, even though adenovirus is a virus.

Mycoplasma pneumoniae told the opposite story in one respect and a striking one in another. Detection of this atypical bacterium was linked to lower odds of a procalcitonin value reaching 0.5 nanograms per milliliter, with an adjusted odds ratio of just 0.22 and a confidence interval of 0.13 to 0.37. Yet the same pathogen was the strongest predictor of a prolonged hospitalization in the entire study: children with Mycoplasma pneumoniae detections had 2.57 times the odds of staying seven days or longer, with a confidence interval of 1.99 to 3.31. The association with long stays persisted when the analysis was restricted to children sampled early, yielding adjusted odds ratios of 2.50 in the 24-hour subset and 2.76 in the 48-hour subset, suggesting that the finding was not an artifact of delayed testing after several days of illness.

Multiple detections carried their own signal. Children in whom two or more targets were identified had 2.21 times the odds of a hospital stay of at least seven days compared with panel-negative children, with a confidence interval of 1.73 to 2.82. Co-detection is increasingly recognized as common rather than exceptional in pediatric respiratory medicine, and this result adds to the evidence that a positive panel with more than one organism is not simply noise. Whether co-detected pathogens interact biologically, or whether multiple detections mark a different underlying illness severity, remains an open question that this observational design cannot resolve.

The authors are careful, and rightly so, about what these numbers do and do not mean. A retrospective cohort of routine clinical data cannot establish that a pathogen causes a longer stay or a higher biomarker value. Unmeasured bacterial infection, variation in treatment, differences in illness severity at presentation, and the clinical selection of which children get tested all could shape the observed associations. The study also relied on routinely recorded biomarkers rather than protocol-mandated sampling, and chest radiograph findings were classified from reports in a way the authors flag as unvalidated. These limitations are acknowledged explicitly in the paper, and the sensitivity analyses, while reassuring, do not convert association into causation.

Even with those caveats, the clinical implications are worth pondering. The adenovirus finding serves as a caution against over-relying on C-reactive protein and procalcitonin as bacterial discriminators: a vigorous inflammatory response in a child with a positive adenovirus swab may reflect the virus itself rather than a bacterial co-infection demanding antibiotics. Conversely, the Mycoplasma pneumoniae result suggests that a low procalcitonin does not guarantee a short or uncomplicated course. Mycoplasma infections in children are notorious for protracted coughing illness and, in some regions, rising macrolide resistance, and a biomarker pattern that looks reassuringly quiet may coexist with a hospitalization that stretches well past a week.

For researchers, the study illustrates the value of linking routine diagnostic panels to outcomes data at scale, and of stress-testing associations with early-sampling restrictions and timing reconciliation before publication. The supplementary materials document an unusually thorough audit trail, including chart-reviewed corrections of PCR timing records and analyses omitting individually corrected entries. For clinicians, the takeaway is more modest but useful: pathogen identity, inflammatory biomarkers, and hospital course each carry independent information, and none substitutes for the others. As multiplex panels become faster and cheaper, studies like this one help map which detections actually matter for the questions that keep a child in the hospital, and which are simply molecular footnotes to an illness that will resolve on its own schedule.

Subject of Research: Associations between multiplex respiratory pathogen detections, inflammatory biomarkers, and length of stay in children hospitalized with acute respiratory tract infections

Article Title: Respiratory panel detections, inflammatory biomarkers, and length of stay in children hospitalized with acute respiratory tract infections: a retrospective cohort study

Article References: Sun, Y., Li, S., Wang, Y., Zhang, S., Teng, S., & Qi, Z. (2026). Respiratory panel detections, inflammatory biomarkers, and length of stay in children hospitalized with acute respiratory tract infections: a retrospective cohort study. BMC Infectious Diseases. https://doi.org/10.1186/s12879-026-14559-x

Image Credits: AI Generated

DOI: 10.1186/s12879-026-14559-x

Keywords: acute respiratory tract infection, children, multiplex PCR, human adenovirus, Mycoplasma pneumoniae, procalcitonin, C-reactive protein, length of stay, biomarkers, pediatrics, retrospective cohort, hospitalization