Abstract
Background:
Community-acquired pneumonia (CAP) is a major contributor to the high burden of acute lower respiratory tract infections in children, characterized by significant incidence and mortality. While targeted next-generation sequencing (tNGS) provides a rapid, accurate, and comprehensive approach to pathogen identification, potentially enabling timely and precise clinical decisions, its application in pediatric populations is not well established. Therefore, this study was designed to investigate the clinical utility of tNGS for detecting pathogens in children with CAP.
Methods:
This retrospective study enrolled 838 children diagnosed with CAP, who were admitted to Chengdu Women and Children’s Central Hospital and underwent tNGS testing between February 2023 and May 2025. Additionally, the results of conventional microbiological tests were collected for comparative analysis to characterize the respiratory pathogen profile in this pediatric cohort.
Results:
The effective detection rate of tNGS (94.0% 788/838) was significantly higher than that of conventional microbiological tests (CMTs) (71.0% 483/680), and the combination of both methods increased the rate to 97.3% (815/838) (P < 0.0001). In contrast to its superior bacterial and atypical pathogen detection, tNGS showed a lower viral detection rate than CMTs. A high positive agreement between the two methods was observed for Mycoplasma pneumoniae (MP) and Haemophilus influenzae, whereas a high negative agreement was found for viral infections. MP was the most frequently detected pathogen overall, followed by Haemophilus influenzae and Streptococcus pneumoniae among bacteria, and rhinovirus and respiratory syncytial virus (RSV) among viruses. Age-specific analysis revealed distinct patterns: viral pathogens were more frequently detected in infants, who also showed the highest detection rate of Staphylococcus aureus; preschool children were prone to bacterial infections; and school-aged children were predominantly affected by atypical pathogens.
Conclusion:
Compared to CMTs, tNGS demonstrated superior capability in detecting a broader range of pathogens with higher sensitivity, making it particularly valuable for identifying coinfections. When interpreted alongside clinical manifestations, tNGS provides clinicians with a precise diagnostic tool for the early identification of lower respiratory tract pathogens. This information, when integrated with clinical presentation, can guide more timely and appropriate therapeutic decisions, which may help mitigate disease severity and reduce mortality in children with CAP.
1 Introduction
Lower respiratory tract infections represent a significant health burden in children under five, accounting for substantial morbidity and a mortality rate of 13.3% (). Among these, community-acquired pneumonia (CAP) is a predominant contributor, caused by a wide spectrum of pathogens, including bacteria, viruses, and fungi (; ). Accurate and timely identification of the causative agent is essential for effective clinical management. However, current conventional microbiological tests (CMTs) face critical limitations. Firstly, methods like bacterial culture and serology are time-consuming. Secondly, even molecular methods like qPCR have a restricted detection scope. These shortcomings delay accurate diagnosis, hinder optimal antibiotic stewardship, and may promote antimicrobial resistance, thereby increasing treatment complexity ().
Currently, next-generation sequencing (NGS) technologies—both metagenomic (mNGS) and targeted (tNGS)—represent the forefront of accurate and rapid pathogen detection in respiratory specimens such as sputum and bronchoalveolar lavage fluid (BALF) (). Although mNGS offers comprehensive pathogen coverage and is adept at detecting rare or fastidious organisms, its widespread application is hindered by several factors: high cost, difficulty differentiating colonization from infection, and the absence of standardized interpretation guidelines, especially given the complexity of the respiratory microbiome (; ). In contrast, tNGS, which integrates PCR with NGS, circumvents many of these drawbacks. Notably, studies have shown that tNGS achieves diagnostic performance comparable to mNGS for respiratory pathogens, may be superior for DNA virus detection, and provides this at a substantially lower cost while still covering up to 95% of respiratory infection etiologies (; ). Moreover, tNGS demonstrates high concordance with both clinical diagnoses and conventional test results, and its resistance gene detection correlates well with phenotypic antimicrobial susceptibility (). Nevertheless, evidence regarding the application of tNGS specifically for pediatric CAP remains scarce, and its clinical utility in this population requires further validation.
This retrospective study analyzed pathogen testing data from 838 children who underwent both tNGS and CMTs. We aimed to compare the detection rates of these two methods for pulmonary pathogens, evaluate the diagnostic utility of tNGS in pediatric CAP, and characterize the respiratory microbial profile in this cohort.
2 Materials and methods
2.1 Data source and ethical statement
We retrospectively analyzed data from 838 children hospitalized with CAP at Chengdu Women and Children’s Central Hospital from February 2023 to May 2025, all of whom underwent tNGS testing on BALF or sputum samples. Demographic information, clinical data, and results from conventional microbiological tests (including respiratory virus qPCR, sputum bacterial culture, and IgM antibody detection) were collected. Bronchoscopy was performed by experienced clinicians following safety protocols, with no serious adverse events reported. Samples were sent to a third-party laboratory for tNGS analysis. The study was approved by the Institutional Ethics Committee of Chengdu Women and Children’s Central Hospital and complied with the Declaration of Helsinki. Written informed consent for procedures was obtained from legal guardians; however, due to the retrospective and anonymized nature of the data analysis, the ethics committee waived the requirement for additional informed consent for this study.
2.2 Inclusion criteria
1. The diagnosis complied with the Chinese guidelines for the diagnosis and treatment of community-acquired pneumonia in children (2023 version); 2. During hospitalization, undergo tNGS testing or CMTs testing in BALF or sputum; 3. Age < 18 years old; 4. Child has complete demographic and clinical data, with no more than 20% of relevant data missing.
2.3 Exclusion criteria
1. There are congenital or secondary immune deficiencies, genetic metabolic diseases, severe hematological diseases and other diseases; 2. Have a history of taking immunomodulators; 3. Incomplete medical history information; 4. Refuse to conduct tNGS and CMTs tests.
2.4 Criteria for judging SCAP
Severe community-acquired pneumonia (SCAP) in children was defined as CAP presenting with any of the following: (1) severe respiratory distress (e.g., shortness of breath, persistent hypoxemia); (2) systemic toxic symptoms (e.g., persistent high fever, consciousness disorder); (3) major complications (e.g., heart failure, shock); or (4) radiographic findings of extensive pulmonary involvement, including large-area consolidation in one or both lobes, atelectasis, or moderate-to-large pleural effusion.
2.5 tNGS detection
BALF or sputum samples were collected for tNGS testing; For infants unable to expectorate, samples were obtained via nasopharyngeal suction. All samples were sent to a third-party laboratory (Chengdu Huayin Medical Laboratory) for tNGS analysis.
2.5.1 Sample processing and nucleic acid extraction
Upon receipt, 600 µL of each sample was lysed using glass beads (0.5 mm) in a TissueLyser II (Qiagen, Hilden, Germany) at 30 Hz for 10 minutes. DNA and RNA were co-extracted using the TIANGEN Nucleic Acid Extraction Kit (TIANGEN Biotech, Beijing, China) according to the manufacturer’s instructions. Extracted nucleic acids were quantified by Qubit 4.0 (Thermo Fisher Scientific, Waltham, MA, USA) and stored at -80 °C until use.
2.5.2 Library preparation and ultra-multiplex PCR
A targeted ultra-multiplex PCR assay was designed to simultaneously amplify 125 clinically common respiratory pathogens (including 68 bacteria, 42 viruses, 11 fungi, and 4 atypical pathogens). The primer pool was designed to target highly conserved regions of each pathogen’s genome, with amplicon sizes ranging from 150 to 300 bp. Multiplex PCR was performed using the Multiplex PCR Kit (Qiagen) with 20 ng of input DNA/cDNA under the following conditions: 95 °C for 15 min; 35 cycles of 94 °C for 30 s, 58 °C for 90 s, and 72 °C for 60 s; and a final extension at 72 °C for 10 min. PCR products were purified using AMPure XP beads (Beckman Coulter, Brea, CA, USA).
2.5.3 Sequencing and bioinformatics
Libraries were sequenced on an Illumina MiniSeq platform (Illumina, San Diego, CA, USA) using 2 × 150 bp paired-end reads. A minimum of 2 million raw reads per sample was required. Raw sequencing data were demultiplexed and adapter-trimmed using fastp (v0.23.2). Human host sequences were removed by alignment to the human reference genome hg38 (GRCh38.p14) using bowtie2 (v2.4.5). Remaining reads were aligned to a curated reference database (RefSeq, NCBI, April 2024 release) containing complete genomes of the 125 target pathogens. Taxonomic classification was performed using Kraken2 (v2.1.2) with a confidence threshold of 0.7. Pathogen reads were normalized to reads per million (RPM) after host subtraction.
2.5.4 Limit of detection and reporting threshold
The analytical limit of detection (LoD) of the assay, determined using quantified control plasmids spiked into negative BALF matrix, was 50 copies/mL for bacterial targets and 200 copies/mL for viral targets. However, to distinguish true infection from colonization or background contamination, a clinical reporting threshold of >1×10³ copies/mL was applied. This threshold was empirically established based on preliminary data from 50 pilot pediatric CAP samples and consensus among three senior clinicians (see section 2.7 for clinical adjudication process).
2.6 CMTs
Conventional microbiological tests included qPCR on nasopharyngeal swabs (for RSV, MP, adenovirus, human rhinovirus, influenza A virus, and influenza B virus), sputum culture, and serological testing. Sputum samples were collected prior to antibiotic administration—by expectoration into sterile tubes for older children or via nasopharyngeal suction for infants and toddlers, and cultured for 72 hours. Sputum culture was attempted in all patients but was successful (produced adequate specimen) in 680 cases, and 18 sputum samples were specifically submitted for tNGS.
Serum samples from a subset of patients were tested for IgM antibodies against common respiratory pathogens (Legionella pneumophila, MP, Coxiella burnetii, C. pneumoniae, adenovirus, RSV, influenza A and B viruses, and parainfluenza virus).
2.7 tNGS result determination
Due to the current lack of standardized interpretation criteria for tNGS, a stringent criterion of >1×10³ copies/mL was established to distinguish true pathogens from colonizing flora, thereby mitigating the influence of sequencing depth and background microbial load. All tNGS-positive results were then independently evaluated by two experienced pediatricians, who integrated clinical manifestations, conventional test results, and radiographic findings according to pediatric CAP guidelines. Any diagnostic discrepancy was resolved by consultation with a third senior clinician, ensuring the reliability of the final pathogen assignment.
3 Statistical methods
Data analysis was conducted using SPSS27.0 software, and P< 0.05 was considered statistically significant. Categorical data were expressed as frequency counts and percentages. Measurement data were expressed as mean ± standard deviation (x ± SD). The significance of categorical data was tested by chi-square test or Fisher’s exact test, and the consistency of the two detection methods was evaluated by Kappa coefficient (). Due to multiple comparisons, Bonferroni correction was applied where appropriate. Adjusted significance thresholds are noted in the relevant tables.
4 Result
4.1 Research population
A total of 838 children diagnosed with CAP and meeting the inclusion criteria were enrolled in this study. All patients underwent tNGS testing on lower respiratory tract specimens, including 820 BALF samples and 18 sputum samples. The cohort consisted of 436 males and 402 females, with a male-to-female ratio of 1.08:1. The mean age was 5.27 ± 3.36 years. Age distribution was as follows: 104 infants (< 1 year), 101 toddlers (1–3 years), 264 preschool children (3–6 years), and 369 school-aged children (> 6 years). Based on disease severity, 635 patients were classified as having SCAP and 203 as having non-severe CAP.
The number of patients who underwent each conventional test varied due to sample availability and clinical indications: qPCR was performed in 752 patients, sputum culture was attempted in all patients but yielded adequate specimens in 680 patients, and IgM antibody testing was performed in a subset of 19 patients based on clinical suspicion.
4.2 tNGS test results
tNGS demonstrated a total detection rate of 99.9% (837/838) and an effective detection rate of 94.0% (788/838) in this pediatric cohort. Total detection rate (99.9%) refers to any tNGS signal above the analytical limit of detection (50 copies/mL). Effective detection rate (94.0%) required the clinical reporting threshold (>1×10³ copies/mL) and independent clinical adjudication to exclude likely colonization or contamination. From the 788 effective cases, 42 pathogens were identified, including 20 bacteria, 16 viruses, 4 fungi, and 2 atypical pathogens (MP and C. pneumoniae). The most prevalent pathogen was MP (36.4%, 305/838), followed by Haemophilus influenzae (20.2%, 169/838), Streptococcus pneumoniae (19.9%, 167/838), rhinovirus (8.7%, 73/838), C. pneumoniae (8.0%, 67/838), and Streptococcus intermedius (8.0%, 67/838). Other detected organisms with frequencies ≥3% included Propionibacterium acnes (7.3%), parainfluenza virus (6.4%), Staphylococcus aureus (6.1%), cytomegalovirus (CMV) (6.0%), Moraxella catarrhalis (5.0%), RSV (4.8%), Epstein-Barr virus (EBV) (4.7%), human herpesvirus (4.4%), and adenovirus (3.5%). The complete pathogen profile, including less frequently detected microorganisms, is presented in Figure 1.
Figure 1
4.3 CMTs results
CMTs achieved an overall effective detection rate of 71.0% (483/680), with method-specific rates as follows: qPCR 58.9% (443/752), sputum culture 19.6% (133/680), and IgM antibody testing 47.4% (9/19). A total of 18 pathogens were identified by CMTs, including 11 bacteria, 6 viruses, and MP. MP was the most frequently detected pathogen (32.3%, 243/752), followed by rhinovirus (20.1%, 151/752), Haemophilus influenzae (7.1%, 48/680), Streptococcus pneumoniae (5.6%, 38/680), adenovirus (5.9%, 44/752), and RSV (5.7%, 43/752) (Figure 2). When combined, tNGS and CMTs increased the overall detection rate to 97.3% (815/838). Importantly, the effective detection rate of tNGS (94.0%, 788/838) was significantly higher than that of CMTs (71.0%, 483/680) (χ² = 293.141, P = 2.2141E-64) (Figure 3).
Figure 2
Figure 3
4.4 Consistency comparison between tNGS and CMTs
Using qPCR and sputum culture results as reference standards, we calculated the positive and negative percent agreement between tNGS and CMTs for eight common respiratory pathogens. Among these, MP demonstrated the highest positive agreement, followed by Haemophilus influenzae and Streptococcus pneumoniae. In contrast, positive agreement for viral pathogens was generally low. The detailed concordance between tNGS and CMTs detection is presented in Table 1.
Table 1
| Pathogens | Positive percent agreement | Negative percent agreement | Kappa |
|---|---|---|---|
| MP | 89.30% (217/243) | 87.43% (445/509) | 0.737 |
| rhinovirus | 27.15% (41/151) | 96.01% (577/601) | 0.294 |
| Haemophilus influenza | 64.58% (31/48) | 83.07% (525/632) | 0.255 |
| adenovirus | 45.45% (20/44) | 98.87% (700/708) | 0.534 |
| RSV | 48.84% (21/43) | 98.31% (697/709) | 0.529 |
| Streptococcus pneumoniae | 50% (19/38) | 82.71% (531/642) | 0.153 |
| Moraxella catarrhalis | 35% (7/20) | 96.21% (635/660) | 0.242 |
| Influenza A virus | 28.57% (4/14) | 99.73% (736/738) | 0.393 |
Consistency comparison of tNGS and CMTs detection for common respiratory pathogens in children.
Identical letters indicate no significant difference (p > 0.05), whereas different letters indicate a significant difference (p < 0.05).
Kappa coefficients were interpreted as: <0.00 poor, 0.00–0.20 slight, 0.21–0.40 fair, 0.41–0.60 moderate, 0.61–0.80 substantial, 0.81–1.00 almost perfect. Accordingly, MP showed substantial agreement (κ=0.737), adenovirus and RSV moderate agreement (κ=0.534 and 0.529), and other pathogens fair to slight agreement.
4.5 Pathogen co-occurrence analysis
To explore the frequency of pathogen co-detection, we constructed a co-occurrence heatmap for the eleven most prevalent pathogens (Supplementary Figure 1). Among the 788 tNGS-positive cases, 412 (52.28%) had a single pathogen detected, 224 (28.43%) had two pathogens, and 152 (19.29%) had three or more. The most frequent co-occurrence pair was Streptococcus pneumoniae with Haemophilus influenza (n=60), followed by Haemophilus influenza with MP (n=33).
4.6 Epidemiological characteristics of respiratory pathogens
Epidemiological analysis was performed for 11 common respiratory microorganisms, examining their overall detection rates, as well as age and gender susceptibility patterns. As shown in Figure 4, MP exhibited the highest total detection rate at 39.50% (331/838), followed by Haemophilus influenzae and Streptococcus pneumoniae, both at 22.20% (186/838). Rhinovirus ranked next with a detection rate of 21.84% (183/838). Other detected pathogens included C. pneumoniae (8.00%, 67/838), RSV (7.40%, 62/838), Staphylococcus aureus (6.80%, 57/838), Moraxella catarrhalis (6.56%, 55/838), parainfluenza virus (6.44%, 54/838), adenovirus (6.21%, 52/838), and CMV (5.97%, 50/838).
Figure 4
Analysis of demographic susceptibility patterns revealed that male children were more susceptible to CMV (x2 = 4.165, P = 0.041) and Streptococcus pneumoniae (x2 = 4.139, P = 0.042), with no significant gender differences for other pathogens. But after Bonferroni correction for 11 comparisons (adjusted P = 0.0045), the nominally significant differences observed for CMV and Streptococcus pneumoniae were no longer statistically significant. Age-stratified analysis showed distinct profiles: school-aged children were predominantly affected by atypical pathogens (MP and C. pneumoniae); preschool children by Haemophilus influenzae and Moraxella catarrhalis; and infants and toddlers by RSV, parainfluenza virus, and CMV. Notably, after Bonferroni correction for 11 comparisons, adenovirus, rhinovirus and Staphylococcus aureus showed no significant differences across age groups (Table 2).
Table 2
| Pathogens | Infant positivity rate | Toddler positivity rate | Preschool children positivity rate | School-age children positivity rate | X2 | P value (Bonferroni correction) |
|---|---|---|---|---|---|---|
| MP | 11.54% (12/104) | 23.76% (24/101) | 36.74% (97/264) | 53.66% (198/369) | 76.287 | 1.9193E-16 |
| Streptococcus pneumoniae | 17.31% (18/104) | 32.67% (33/101) | 27.27% (72/264) | 17.07% (63/369) | 17.407 | 0.000583 |
| Haemophilus influenza | 25.00% (26/104) | 21.78% (22/101) | 29.92% (79/264) | 15.99% (59/369) | 17.846 | 0.000473 |
| Moraxella catarrhalis | 8.65% (9/104) | 8.91% (9/101) | 9.85% (26/264) | 2.98% (11/369) | 14.017 | 0.002883 |
| C. pneumoniae | 0.96% (1/104) | 1.98% (2/101) | 3.41% (9/264) | 14.91% (55/369) | 43.462 | 1.9633E-9 |
| RSV | 18.27% (19/104) | 12.87% (13/101) | 6.44% (17/264) | 3.52% (13/369) | 30.798 | 9.3767E-7 |
| Parainfluenza virus | 17.31% (18/104) | 12.87% (13/101) | 4.17% (11/264) | 3.25% (12/369) | 35.788 | 8.3042E-8 |
| adenovirus | 3.85% (4/104) | 11.88% (12/101) | 7.20% (19/264) | 4.61% (17/369) | 8.651 | 0.034316 |
| CMV | 29.81% (31/104) | 4.95% (5/101) | 3.03% (8/264) | 1.63% (6/369) | 121.994 | 2.87E-26 |
| Staphylococcus aureus | 12.50% (13/104) | 9.90% (10/101) | 5.68% (15/264) | 5.15% (19/369) | 8.970 | 0.029698 |
| rhinovirus | 28.85% (30/104) | 26.73% (27/101) | 22.73% (60/264) | 17.89% (66/369) | 7.909 | 0.047940 |
Age distribution of different respiratory pathogens in children with CAP.
Bonferroni correction was applied for the 11 age-group comparisons (one per pathogen). Statistical significance was defined as P < 0.05/11 = 0.0045. P-values below 0.0045 are considered significant after correction.
5 Discussion
CAP remains a leading cause of morbidity and mortality in children worldwide, underscoring the critical need for rapid and accurate pathogen identification to guide timely clinical interventions and improve patient outcomes (). In this study, we evaluated the diagnostic performance of tNGS for respiratory pathogen detection in pediatric CAP and compared it with CMTs. Our findings demonstrate that tNGS achieved a significantly higher detection rate than CMTs (P < 0.05), consistent with previous reports (). Notably, MP emerged as the predominant pathogen in both detection platforms, with detection rates exceeding 30%. Furthermore, the positive and negative percent agreement between tNGS and CMTs for MP detection reached approximately 80%, a finding that diverges from some earlier studies (). This discrepancy may be attributed to the global MP pandemic observed since 2023, which has been particularly pronounced in East Asia. In China, MP infection rates have surged to as high as 47.89% (). The resultant high pathogen load in infected children likely minimizes methodological variability, potentially explaining the comparable detection performance between tNGS and CMTs observed in our cohort.
However, for other common pathogens, detection rates varied significantly between methods. Notably, CMTs demonstrated superior detection rates for respiratory viruses compared to tNGS, with tNGS showing positive agreement below 50% but negative agreement exceeding 96% for viral pathogens. This indicates that CMTs remain the preferred method for rapid viral detection in clinical settings. Nevertheless, tNGS proved valuable for detecting viruses in CMTs-negative samples, thereby increasing overall viral detection rates and reducing false negatives. Consistent with previous reports (), we found that tNGS-assisted detection achieved a positivity rate of up to 81.1% in CMTs-negative samples. This high false-negative rate with CMTs may be attributed to sampling timing relative to infection dynamics. Studies on SARS-CoV-2 have demonstrated that viral load peaks in the upper respiratory tract during early symptomatic phases but declines more rapidly in the upper tract than in the lower tract as infection progresses, particularly in severe cases (; ). Since qPCR typically uses nasopharyngeal swabs, CMTs are most effective during early infection. While CMTs offer advantages in turnaround time and cost, their limited sensitivity for low viral loads and mixed infections renders them suitable primarily for rapid diagnosis of common respiratory viruses in appropriate clinical contexts (). For patients with negative CMTs results but persistent high fever, suspected severe disease, or concern for atypical viral infections, tNGS should be considered as a complementary tool to enhance viral detection. Indeed, our study demonstrated that combining CMTs with tNGS increased the overall effective detection rate to 97.3%. From a logistical perspective, the turnaround time for tNGS in our study was approximately 48–72 hours from sample receipt to result report, which is longer than qPCR (4–6 hours) but comparable to bacterial culture (48–72 hours). The cost per tNGS sample was approximately 80 USD, which is higher than individual qPCR assays (15–25 per target) but substantially lower than mNGS (300–500 USD). Therefore, while tNGS may not replace qPCR for rapid point-of-care testing, it offers a cost-effective solution when a broad differential diagnosis or co-infection assessment is needed.
tNGS demonstrated superior bacterial detection compared to CMTs, particularly for Haemophilus influenzae (positive agreement 64.58%, negative agreement 83.07%) and Streptococcus pneumoniae (50% and 82.71%, respectively), indicating its ability to identify pathogens missed by sputum culture. While tNGS offers advantages in turnaround time, detection breadth, and identification of mixed infections—including fastidious organisms like Mycobacterium tuberculosis (; )—results must be interpreted cautiously due to potential false positives from background contamination or colonization (). In our study, when tNGS and CMTs results disagreed, we did not automatically consider tNGS as correct. Instead, we used a clinical adjudication panel that reviewed each case’s complete clinical presentation, radiographic findings, and treatment response. For example, a tNGS-positive but CMTs-negative result for Streptococcus pneumoniae was classified as a true positive if the child had lobar consolidation and responded definitively to beta-lactam therapy; if the child was afebrile, had no radiographic infiltrate, and recovered without antibiotics, the tNGS result was considered likely colonization or contamination. This approach minimized misclassification bias but also highlighted the absence of a true gold standard—a limitation we address below.
Beyond pathogen detection, tNGS revealed respiratory microbial community characteristics, with Streptococcus intermedius and Propionibacterium species contributing to oral microbial colonization in early childhood (). In the post-pandemic era, MP remains the predominant pathogen, followed by H. influenzae, S. pneumoniae, rhinovirus, and RSV, consistent with previous reports (). MP exhibits 3–7 years epidemic cycles, influencing detection rates (). Age-stratified analysis revealed distinct susceptibility patterns: infants and toddlers were more susceptible to viral infections; preschool children to bacterial infections; and school-aged children to atypical pathogens. These patterns reflect age-dependent factors including immunological maturation, anatomical development, and social exposure patterns (; ; ). In our study, infants showed the highest CMV detection rates among all age groups. This pattern is consistent with established understanding that postnatal CMV acquisition in infancy occurs primarily through breast milk from seropositive mothers, a transmission route well documented in the literature (). All CMV-infected infants were asymptomatic with normal hearing, consistent with previous studies ().
This study has several limitations. First, as a single-center study conducted in Chengdu, the findings may primarily reflect the local epidemiological characteristics of respiratory pathogens and may not be generalizable to the broader pediatric population. Second, the absence of a true gold standard for respiratory pathogen detection precluded the calculation of accurate sensitivity and specificity for either tNGS or CMTs; therefore, we could only assess the performance of tNGS through concordance analysis with conventional methods. Future studies addressing this limitation could employ latent class models or Bayesian methods to estimate sensitivity and specificity in the absence of a gold standard (; ). Third, the lack of standardized criteria for tNGS result interpretation may have led to underestimation of its true diagnostic performance, particularly for low-abundance microorganisms or those not detected by CMTs. Fourth, fungal pathogens were not included in the CMTs detection panel, preventing direct comparison of the two methods for fungal detection. Consequently, our findings should be interpreted with caution.
Future multicenter studies should address the following: (1) standardize tNGS interpretation criteria by establishing pathogen-specific quantitative thresholds based on bronchoalveolar lavage fluid from healthy controls; (2) incorporate longitudinal follow-up (e.g., 30-day outcomes) to assess the impact of tNGS results on antibiotic de-escalation, length of hospital stay, and mortality; (3) compare tNGS-guided therapy versus CMTs-guided therapy in a randomized controlled design. Such studies might reveal whether routine tNGS use reduces inappropriate antibiotic use by 20-30% and shortens hospitalization by 2–3 days in children with severe CAP.
6 Conclusion
In summary, while tNGS demonstrates superior diagnostic performance compared to CMTs for pediatric CAP, reliance on a single testing modality is insufficient for accurate etiological diagnosis. The optimal diagnostic strategy involves a rational selection of detection methods—guided by the stage of illness, clinical characteristics, and initial auxiliary findings—to effectively identify viral, bacterial, and atypical pathogens. Integrating tNGS results with clinical manifestations significantly enhances diagnostic accuracy. Notably, for children with SCAP, the combined use of CMTs and tNGS markedly improves pathogen detection, enabling timely therapeutic adjustments that can improve prognosis and reduce mortality.
Statements
Data availability statement
The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.
Ethics statement
The studies involving humans were approved by the Institutional Ethics Committee of Chengdu Women and Children’s Central Hospital. The studies were conducted in accordance with the local legislation and institutional requirements. Written informed consent for participation was not required from the participants or the participants’ legal guardians/next of kin in accordance with the national legislation and institutional requirements.
Author contributions
LMY: Writing – original draft, Writing – review & editing. HZ: Writing – review & editing. HF: Writing – review & editing. HD: Software, Writing – original draft, Writing – review & editing. JC: Writing – original draft, Writing – review & editing. LY: Writing – original draft, Writing – review & editing. LZ: Conceptualization, Investigation, Writing – original draft, Writing – review & editing.
Funding
The author(s) declared that financial support was received for this work and/or its publication. This work was supported by Sichuan Province Maternal and Child Medical Science and Technology Innovation Project (2024FX03) and Chengdu City Key Research and Development Project (2024-YF05-00749-SN).
Acknowledgments
The authors would like to thank the Department of School of Medicine, University of Electronic Science and Technology of China, Chengdu, and the Chengdu Women’s and Children’s Central Hospital for providing technical supports for this study.
Conflict of interest
The author(s) declared that this work was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
Generative AI statement
The author(s) declared that generative AI was not used in the creation of this manuscript.
Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.
Publisher’s note
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.
Supplementary material
The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fcimb.2026.1827409/full#supplementary-material
References
1
ChenP. Z.BobrovitzN.PremjiZ. A.KoopmansM.FismanD. N.GuF. X. (2021). SARS-CoV-2 shedding dynamics across the respiratory tract, sex, and disease severity for adult and pediatric COVID-19. eLife10, e70458. doi: 10.7554/eLife.70458
2
ChenB.GaoL.ChuQ.ZhouT.TongY.HanN.et al. (2024). The epidemic characteristics of Mycoplasma pneumoniae infection among children in Anhui, China 2015–2023. Microbiol. Spectr.12(10), e0065124. doi: 10.1128/spectrum.00651-24
3
DendukuriN.JosephL. (2001). Bayesian approaches to modeling the conditional dependence between multiple diagnostic tests. Biometrics57, 158–167. doi: 10.1111/j.0006-341x.2001.00158.x
4
DeshmukhH.WhitsettJ.ZachariasW.WayS. S.MartinezF. D.MizgerdJ.et al. (2025). Impact of viral lower respiratory tract infection (LRTI) in early childhood (0-2 years) on lung growth and development and lifelong trajectories of pulmonary health: A National Institutes of Health (NIH) workshop summary. Pediatr. Pulmonol.60, e27357. doi: 10.1002/ppul.27357
5
DiaoZ.HanD.ZhangR.LiJ. (2022). Metagenomics next-generation sequencing tests take the stage in the diagnosis of lower respiratory tract infections. J. Adv. Res.38, 201–212. doi: 10.1016/j.jare.2021.09.012
6
FalzoneL.MussoN.GattusoG.BongiornoD.PalermoC. I.ScaliaG.et al. (2020). Sensitivity assessment of droplet digital PCR for SARS-CoV-2 detection. Int. J. Mol. Med.46, 957–964. doi: 10.3892/ijmm.2020.4673
7
GBD 2019 LRI Collaborators (2022). Age-sex differences in the global burden of lower respiratory infections and risk factors 1990-2019: results from the global burden of disease study 2019. Lancet Infect. Dis.22, 1626–1647. doi: 10.1016/S1473-3099(22)00510-2
8
GBD 2019 Under-5 Mortality Collaborators (2021). Global, regional, and national progress towards Sustainable Development Goal 3.2 for neonatal and child health: all-cause and cause-specific mortality findings from the Global Burden of Disease Study 2019. Lancet398, 870–905. doi: 10.1016/S0140-6736(21)01207-1
9
GuZ.LiuT.LiJ.SongC.WangX.TangY.et al. (2025). Enhancing upper respiratory tract infection detection: exploring qPCR negative respiratory samples using targeted next-generation sequencing. Infection53, 2433–2442. doi: 10.1007/s15010-025-02578-w
10
HandelmanC. S.OsborneG. (1976). Growth of the nasopharynx and adenoid development from one to eighteeen years. Angle Orthod.46, 243–259. doi: 10.1043/0003-3219(1976)046<0243:GOTNAA>2.0.CO;2
11
HeS.XueW.WuX.LiangZ.GaoJ.QinW.et al. (2025). The application value of targeted next-generation sequencing using bronchoalveolar lavage fluid samples in community-acquired pneumonia in children. J. Infect. Chemother. Off. J. Jpn. Soc Chemother.31, 102610. doi: 10.1016/j.jiac.2025.102610
12
HuiS. L.ZhouX. H. (1998). Evaluation of diagnostic tests without gold standards. Stat. Methods Med. Res.7, 354–370. doi: 10.1177/096228029800700404
13
JinW.WangM.WangY.ZhuB.WangQ.ZhouC.et al. (2025). Targeted next-generation sequencing: a promising approach for Mycobacterium tuberculosis detection and drug resistance when applied in paucibacillary clinical samples. Microbiol. Spectr.13, e0312724. doi: 10.1128/spectrum.03127-24
14
JohnsonK. E.Hernandez-AlvaradoN.BlackstadM.HeiselT.AllertM.FieldsD. A.et al. (2024). Human cytomegalovirus in breast milk is associated with milk composition and the infant gut microbiome and growth. Nat. Commun.15, 6216. doi: 10.1038/s41467-024-50282-4
15
KennedyB.PeuraS.HammarU.VicenziS.HedmanA.AlmqvistC.et al. (2019). Oral microbiota development in early childhood. Sci. Rep.9, 19025. doi: 10.1038/s41598-019-54702-0
16
LanzieriT. M.ChungW.FloresM.BlumP.CavinessA. C.BialekS. R.et al. (2017). Hearing loss in children with asymptomatic congenital cytomegalovirus infection. Pediatrics139, e20162610. doi: 10.1542/peds.2016-2610
17
LeeK.-L.LeeC.-M.YangT.-L.YenT.-Y.ChangL.-Y.ChenJ.-M.et al. (2021). Severe Mycoplasma pneumoniae pneumonia requiring intensive care in children 2010-2019. J. Formos. Med. Assoc. Taiwan. Yi Zhi120, 281–291. doi: 10.1016/j.jfma.2020.08.018
18
LiS.TongJ.LiuY.ShenW.HuP. (2022). Targeted next generation sequencing is comparable with metagenomic next generation sequencing in adults with pneumonia for pathogenic microorganism detection. J. Infect.85, e127–e129. doi: 10.1016/j.jinf.2022.08.022
19
LinX.ZhaiK.LiuB. M.ChenJ. (2025). Metabolism-driven colorimetric “read-to-answer” sensor array for bacterial discrimination and antimicrobial susceptibility testing. Anal. Chem.97, 17040–17049. doi: 10.1021/acs.analchem.5c02762
20
LiuS. S.DengJ. R.WangD. S.GongX.ZhouY. H.GaoX. M. (2016). Nasopharyngeal changes in 8-13 years old healthy children in China: a longitudinal study. Zhonghua Er Bi Yan Hou Tou Jing Wai Ke Za Zhi51, 733–739. doi: 10.3760/cma.j.issn.1673-0860.2016.10.004
21
LiuY.-N.ZhangY.-F.XuQ.QiuY.LuQ.-B.WangT.et al. (2023). Infection and co-infection patterns of community-acquired pneumonia in patients of different ages in China from 2009 to 2020: a national surveillance study. Lancet Microbe4, e330–e339. doi: 10.1016/S2666-5247(23)00031-9
22
McAllisterD. A.LiuL.ShiT.ChuY.ReedC.BurrowsJ.et al. (2019). Global, regional, and national estimates of pneumonia morbidity and mortality in children younger than 5 years between 2000 and 2015: a systematic analysis. Lancet Glob. Health7, e47–e57. doi: 10.1016/S2214-109X(18)30408-X
23
ShangH.ZouS.MaZ.LiangQ.ZhongY.LiL.et al. (2025). Comparative and clinical impact of targeted next-generation sequencing in pediatric pneumonia diagnosis and treatment. Front. Microbiol.16, 1590792. doi: 10.3389/fmicb.2025.1590792
24
SuL. D.ChiuC. Y.GastonD.HoganC. A.MillerS.SimonD. W.et al. (2024). Clinical metagenomic next-generation sequencing for diagnosis of central nervous system infections: Advances and challenges. Mol. Diagn. Ther.28, 513–523. doi: 10.1007/s40291-024-00727-9
25
SunW.ZhengL.KangL.ChenC.WangL.LuL.et al. (2024). Comparative analysis of metagenomic and targeted next-generation sequencing for pathogens diagnosis in bronchoalveolar lavage fluid specimens. Front. Cell. Infect. Microbiol.14, 1451440. doi: 10.3389/fcimb.2024.1451440
26
ToothL. R.OttenbacherK. J. (2004). The kappa statistic in rehabilitation research: an examination. Arch. Phys. Med. Rehabil.85, 1371–1376. doi: 10.1016/j.apmr.2003.12.002
27
TorresA.CillonizC.NiedermanM. S.MenéndezR.ChalmersJ. D.WunderinkR. G.et al. (2021). Pneumonia. Nat. Rev. Dis. Primer7, 25. doi: 10.1038/s41572-021-00259-0
28
WangX.Stelzer-BraidS.ScotchM.RawlinsonW. D. (2022). Detection of respiratory viruses directly from clinical samples using next-generation sequencing: A literature review of recent advances and potential for routine clinical use. Rev. Med. Virol.32, e2375. doi: 10.1002/rmv.2375
29
WölfelR.CormanV. M.GuggemosW.SeilmaierM.ZangeS.MüllerM. A.et al. (2020). Virological assessment of hospitalized patients with COVID-2019. Nature581, 465–469. doi: 10.1038/s41586-020-2196-x
30
XuD.LiS.ChenZ.DuL. (2011). Detection of Mycoplasma pneumoniae in different respiratory specimens. Eur. J. Pediatr.170, 851–858. doi: 10.1007/s00431-010-1360-y
31
ZhangP.LiuB.ZhangS.ChangX.ZhangL.GuD.et al. (2024). Clinical application of targeted next-generation sequencing in severe pneumonia: a retrospective review. Crit. Care28, 225. doi: 10.1186/s13054-024-05009-8
Summary
Keywords
bacterial culture, children, community-acquired pneumonia, targeted next-generation sequencing, virus detection
Citation
Yan L, Zhang H, Fu H, Dong H, Chen J, Yu L and Zhang L (2026) Targeted next-generation sequencing for pediatric community-acquired pneumonia pathogen detection: a single-center study. Front. Cell. Infect. Microbiol. 16:1827409. doi: 10.3389/fcimb.2026.1827409
Received
10 March 2026
Revised
03 June 2026
Accepted
14 July 2026
Published
29 July 2026
Volume
16 - 2026
Edited by
Qing Wei, Wuhan Kindstar Medical Laboratory Co., Ltd., China
Reviewed by
Benjamin M. Liu, George Washington University, United States
Christian Gunning, University of Georgia, United States
Updates
Copyright
© 2026 Yan, Zhang, Fu, Dong, Chen, Yu and Zhang.
This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
*Correspondence: Lei Zhang, 534167313@qq.com
Disclaimer
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article or claim that may be made by its manufacturer is not guaranteed or endorsed by the publisher.