Blog

Introduction

Asthma is a chronic respiratory disease affecting 262 million people worldwide, with children representing a particularly vulnerable population []. High prevalence rates have been reported in urban regions of China and India []; yet, early-life asthma remains both undermanaged in clinical practice and underrepresented in research []. This is due to several challenges, including the fast-paced nature of pediatric outpatient clinics where most children are treated [], the inherent heterogeneity of asthma [], and diagnostic difficulties—especially in children under 6 years old []. These factors contribute to low rates of diagnostic confirmation and inhaled corticosteroid adherence, exacerbating the overall disease burden [,].

To address these issues, the Global Initiative for Asthma (GINA) 2025 report has introduced, for the first time, diagnostic criteria and a definition for suspected asthma in children under 6 years old []. Despite this important step, the characteristics of early-life asthma in outpatient settings remain poorly characterized. This gap is particularly critical since the first 5 years of life constitute a key window for allergen sensitization [], the foremost risk factor for childhood asthma. Therefore, there is an urgent need to establish well-characterized cohorts of young asthmatic children within outpatient settings to improve understanding and management.

To date, 3 primary approaches have been used to study childhood asthma. Large-scale birth cohorts have advanced our understanding of asthma and allergic diseases [-] but require substantial resources and yield limited cases due to disease incidence. Small-scale prospective asthma cohorts often include children with established diagnoses (typically over 6 years old) [-]; yet, this may restrict exploration of early disease heterogeneity. Retrospective cohorts based on hospital records [-] offer real-world insights but are limited by incomplete data and suboptimal biospecimen preservation. Given these limitations, an integrated model that combines prospective and retrospective approaches within routine clinical practice offers a promising alternative.

Emerging strategies enhance the feasibility and rigor of such an integrated model. Real-world evidence methodologies and symptom-driven phenotyping enable early identification of children with asthmatic symptoms, even before formal diagnosis [-]. Standardized electronic medical records (EMRs) improve data completeness and accessibility [,], while electronic patient-reported outcomes (ePROs) capture symptoms and experiences beyond clinical encounters [-]. In addition, residual biospecimens from routine testing can be repurposed for molecular endotyping using high-throughput omics technologies, minimizing invasive procedures [-]. These tools not only reduce cohort heterogeneity but also facilitate data harmonization and future integration, mirroring successful efforts in major birth cohorts [,].

In this study, we developed the Clinical Registry of Childhood Asthma (CRCA)—an outpatient-based integrated research platform that combines symptom-driven recruitment, real-world evidence collection, standardized EMRs, structured ePROs, and residual biospecimens. This framework supports both the establishment of a dynamically updated biorepository and enables phenotyping of early-life asthma. We aimed to characterize children across confirmed, suspected, and excluded asthma categories and identify factors associated with these diagnostic outcomes.

Methods

Study Design

The CRCA is a prospective, longitudinal, digitally enhanced real-world cohort study launched on March 7, 2024, at the Respiratory Clinic of the Children’s Hospital of Chongqing Medical University—a tertiary pediatric referral center in Southwest China. Designed to address the critical gap in well-characterized outpatient cohorts for early-life asthma, the CRCA operates as an open, integrated research framework embedded within routine clinical workflows to enable long-term follow-up (). This manuscript presents a cross-sectional analysis of the baseline data from this ongoing cohort.

‎

This study used a symptom-driven recruitment strategy to proactively identify and enroll children during visits [,]. While this approach may initially include children with a broader spectrum of wheezing illnesses (“off-target” recruitment) [] and is subject to real-world follow-up challenges [-], it is conceptualized not as a limitation but as a fundamental feature enabling the construction of a dynamic, funnel-shaped biorepository for long-term observational research. Through continuous enrollment, the framework accumulates a diverse population, with data collection occurring at scheduled follow-up intervals, which naturally refines over time into a well-characterized cohort of children with confirmed or suspected asthma, while also preserving longitudinal data on the broader natural history of childhood wheezing.

The term “digitally enhanced” refers to our core strategy of synthesizing two complementary data streams collected overtime: (1) hospital-based EMRs for structured clinical data and (2) smartphone-based ePROs for repeatedly capturing symptom and experience data from caregivers beyond clinical encounters. This dual-stream approach ensures comprehensive and continuous longitudinal data acquisition within a high-volume outpatient setting.

The primary outcomes for this article are the detailed cross-sectional characteristics of children with early-life asthma, stratified by diagnostic status (confirmed, suspected, or excluded). Secondary outcomes include metrics related to the implementation of the CRCA study, including the feasibility (>70% ePRO completion and biospecimens collection rates) and sustainability of recruitment efforts. In-depth analysis of longitudinal outcomes is beyond the scope of this initial report, but it is a core purpose of the ongoing cohort. Reporting follows the STROBE (Strengthening the Reporting of Observational Studies in Epidemiology) guidelines ().

Ethical Considerations

The study protocol was approved by the Ethics Committee of the Children’s Hospital of Chongqing Medical University (approval number 2024-37) and registered with China’s National Medical Research Registration and Recording Information System (registration number MR-50-24-049642). Written informed consent was obtained from all participating parents, with additional assent from children aged ≥8 years. No financial compensation was provided. The study imposed no additional diagnostic procedures or financial burdens beyond standard care. All procedures adhered to the Declaration of Helsinki, and participant confidentiality was maintained.

Recruitment and Follow-Up

The longitudinal design of CRCA is implemented through a structured recruitment and follow-up protocol. Eligible participants were children under 18 years of age who presented with persistent or recurrent cough, wheezing, or both, lasting at least 1 month. In addition, asthma needed to be suspected by either the referring practitioner or our study pediatricians. Finally, the child’s clinical assessment or a parental request had to indicate the need for blood tests, such as serum-specific IgE (sIgE) testing. Participants who met the inclusion criteria and elected to continue follow-up at the CRCA clinic were enrolled. It is important to note that while the CRCA registry enrolls children across the entire pediatric age range (<18 years) to establish a comprehensive and longitudinal resource, this initial cross-sectional analysis focuses specifically on the early-life period, where diagnostic uncertainty is greatest.

Enrolled participants are followed prospectively every 4-6 months or as clinically indicated, in accordance with GINA-based management []. Study exit was defined as either (1) clinical resolution of symptoms or (2) failure to attend follow-up for over one year. However, given the chronic and relapsing nature of asthma, children who returned to the clinic after study exit, whether for exacerbations or routine management, were re-eligible for data collection ().

Definitions of Asthma Diagnosis

Asthma status was classified for all ages as “confirmed” or “suspected” per GINA 2025 []. For children under 6 years, a “confirmed asthma” diagnosis required meeting 3 criteria. First, the child must have had recurrent acute wheezing episodes, which could occur with or without interval symptoms. Second, alternative diagnoses must have been excluded. Third, a timely response to asthma treatment was necessary. A “suspected asthma” classification was applied when there was a highly suggestive clinical history that lacked objective confirmation. This lack of confirmation could be due to the infeasibility of testing or an incomplete treatment response. In such cases, alternative diagnoses were also considered unlikely. For individuals ≥6 years, “confirmed asthma” required objective evidence of variable expiratory airflow limitation, demonstrated by positive bronchodilator reversibility, excessive peak expiratory flow variability, or a positive bronchial challenge test. “Suspected asthma” was defined by characteristic symptoms in the absence of confirmatory test results, whether due to testing infeasibility or negative or equivocal results despite a suggestive history.

Standard EMRs

The CRCA study implemented a standardized EMR template to ensure consistent data collection. The baseline information module captured comprehensive demographic characteristics, including the child’s full name, date of birth, residential address, contact information, and body weight. The clinical assessment module systematically recorded medical data across multiple domains. These included respiratory symptoms, concomitant conditions, and current medications. It also captured treatment responses, physical examination findings, and laboratory results such as complete blood counts and allergen sensitization profiles (serum sIgE tests). Additionally, objective measures like lung function and fractional exhaled nitric oxide (FeNO) were documented. This module also documented complete medical histories, encompassing perinatal events, personal atopy history, previous severe pneumonia episodes, and mechanical ventilation history. Data were initially collected by research staff during preconsultation assessments, then verified and supplemented during formal consultations and test result reviews ().

‎

ePROs

Parents (primarily mothers) of eligible participants were invited to connect with the official WeChat account of the CRCA study. This account fulfilled 3 essential functions: establishing a direct pediatrician-parent communication channel for clinical inquiries, serving as the platform for ePROs collection, and enabling prioritized scheduling of follow-up outpatient visits.

At enrollment, parents received a voluntary web-based questionnaire administered through WeChat, adapted from our previous and related studies to capture multidimensional quantitative data [-]. This instrument collected information across key domains, including sibling composition, family structure, familial atopy history, residential characteristics, pet ownership, indoor air quality, hygiene practices, household chemical usage patterns, health status, asthma risk factors, pharmacological treatment history, and symptom control metrics using age-appropriate Asthma Control Test (ACT) or Childhood Asthma Control Test (C-ACT). Each questionnaire included a self-rated response quality assessment item to ensure high-quality data collection. The primary analytical objective for this comprehensive environmental exposure data is to determine its associations with key asthma-related outcomes. These outcomes include diagnostic status, symptom control, and exacerbation frequency. Our planned analysis will use multivariate statistical models to quantify these exposure-outcome relationships while adjusting for potential confounders, including demographic and familial risk factors. For longitudinal monitoring, abbreviated follow-up questionnaires collected core demographics alongside age-appropriate ACT or C-ACT measures at each scheduled follow-up time point.

Residual Biospecimens

Residual biospecimens (primarily blood) were collected by CRCA staff through the Clinical Laboratory Department, typically during the afternoon following blood collection. Samples were processed to isolate plasma, serum, peripheral blood mononuclear cells (PBMCs), and other blood cells. Serum and plasma were aliquoted at 200 μL per tube, PBMCs at 500 μL per tube, and other blood cells were stored in one tube per participant. All aliquoted samples were cryopreserved. PBMCs underwent programmed freezing. They were first stored at –80° C for over 12 hours before their final transfer to liquid nitrogen. All samples were ultimately categorized and stored under standardized conditions (–80° C freezers or liquid nitrogen) within our institution’s Biobank Center. All associated metadata were tracked electronically. This process followed protocols established by the Western China Birth Cohort study []. Residual blood specimens were not used in the present analysis to avoid purposeless consumption. However, the scientific value of these biospecimens is well-established. This is demonstrated by prior multiomics applications in Mycoplasma pneumoniae pneumonia research, which used identical specimen collection protocols [].

Quality Control and Data Security

The CRCA working group comprised trained pediatricians and clinical researchers who met monthly to review and optimize operations. All data were stored locally under anonymized identifiers. Access to data and biospecimens required approval from the principal investigator under stringent security protocols.

A dedicated team managed the EMR data extraction. Initial collection and entry were performed by Juan L and HF, and are now managed by PY and JZ. All EMR data were reviewed and verified by EL. ePROs followed rigorous methodological standards from our previous work [,]. Biospecimen handling adhered to stringent Chinese National Standards for human blood biobanking; our group contributed to drafting “Human blood sample biobanking—Part 1: General specifications” (Plan #20253246-T-469) and “Part 2: Peripheral blood mononuclear cells” (Plan #20253247-T-469), publicly accessible on China’s National Public Service Platform for Standards Information. This ensures long-term sample quality and readiness for integration with clinical outcomes.

Statistical Analyses

Data analysis and plotting were performed using R 4.4.1, GraphPad Prism 9, and the BioRender website. Categorical variables are presented as frequencies (percentages) while continuous variables are shown as medians (IQRs) following normality assessment. Missing data were not imputed for 2 primary reasons. First, missingness in lung function and FeNO measurements was deemed not missing at random, as it was highly correlated with disease severity and clinical progression. Second, missing questionnaire data typically involved entire unreturned forms; imputation of such large-scale missingness was considered to introduce unacceptable bias. Therefore, all analyses were performed on complete cases.

Participants were stratified by diagnostic category (confirmed, suspected, and excluded asthma). Comparative analyses applied chi-square tests for categorical variables, and Kruskal-Wallis tests with Bonferroni correction for multiple comparisons. A 2-sided P<.05 was considered statistically significant. For multivariable analyses, three logistic regression models were constructed: (1) Model 1 compared suspected versus excluded asthma; (2) Model 2 compared confirmed versus suspected asthma; and (3) Model 3 compared confirmed versus excluded asthma. We used a tiered variable selection strategy to enhance robustness. Univariate screening was conducted at three thresholds (P<.30, P<.20, and P<.10). The model using the strictest threshold (P<.10) served as the primary final model, while those using more lenient thresholds (P<.30 and P<.20) were used in sensitivity analyses to evaluate the stability of core associations. Within each tier, forward stepwise likelihood ratio selection was applied, with age and sex retained in all models irrespective of significance. All models are based on the cross-sectional baseline diagnosis. The stability of these diagnoses overtime and their evolution will be the subject of future longitudinal analyses of this cohort.

Results

Implementation and Feasibility of the CRCA Study

Between March 2024 and August 2025 (A), we organized 62 weekly outpatient sessions, evaluating an average of 37 patients per session (total 2296 visits; mean consultation time: 7.3 minutes). From this population, 396 eligible patients were enrolled (enrollment rate: 17.2%). Initial and second follow-up assessments were completed by 26.7% (84/315) of eligible patients at the first time point and 43.3% (26/60) of those eligible for the second follow-up, respectively (B). Temporal patterns indicated that follow-up completion rates peaked during school vacations and national holidays, reaching 1.4 times the baseline. These results demonstrate the feasibility of implementing a comprehensive research framework within the constraints of a high-throughput pediatric clinic, and underscore the need for flexible, family-centered follow-up strategies in real-world studies.

The integrated data and biospecimen repository constitutes a core output of the CRCA study. All enrolled children have standardized EMRs, which include 392 complete blood count reports, 389 allergen-sIgE reports, 86 lung function tests, 114 FeNO measurements, 280 parental questionnaires, and 139 ACT or C-ACT assessments. Furthermore, the biorepository has accumulated 2423 residual biospecimens (B), with coverage rates as follows: serum samples from 357 (90.2%) children, plasma from 317 (80.1%), PBMCs from 171 (43.2%), and other blood cell samples from 327 (82.6%), yielding an average coverage of 74.0% across these specimen types. The rate of hemolysis occurrence was 4.0% (16/396). These coverage rates meet the prespecified feasibility indicators for the biorepository component of the study.

‎

Clinical Characteristics and Diagnostic Patterns

The CRCA successfully captured a predominantly young cohort, with 267 (267/396, 67.4%) children younger than 6 years. Patients were classified into 3 diagnostic categories: confirmed asthma (131/396, 33.1%), suspected asthma (179/396, 45.2%), and excluded asthma (86/396, 21.7%; A). Comorbidity profiles differed markedly among groups (B). Allergic rhinitis was most prevalent in the confirmed asthma group (77/131, 58.8%). Children in the excluded asthma group were most frequently diagnosed with chronic cough (64/86, 74.4%) or bronchitis (39/86, 45.3%). In contrast, the suspected asthma group showed a lower prevalence of additional diagnoses, suggesting a clinically distinct, “pure” early wheezing phenotype that is difficult to classify using current criteria.

As shown in , children with suspected or excluded asthma were significantly younger than those with confirmed asthma (P<.001), indicating that the under-6 age group largely comprises uncertain or excluded cases.

‎

These groups also included more first-time visitors (P<.001), highlighting the initial visit as a critical yet uncertain diagnostic window. Significant differences in comorbidity patterns were observed across diagnostic categories. Among children with a previous asthma treatment history (n=250), specialist evaluation confirmed asthma in only 46.8% (117/250) and definitively excluded it in 15.2% (38/250), underscoring substantial community-level under- and overdiagnosis. As expected, objective measures of type 2 inflammation and lung function—including aeroallergen sensitization, blood eosinophil counts, and FeNO—differed significantly across groups (P<.001), supporting the clinical validity of the diagnostic stratification.

Self-Reported Characteristics and Patient-Physician Discordance

The study achieved a 70.7% (280/396) questionnaire completion rate, with no significant demographic or diagnostic differences between responders and nonresponders. A central finding was a pronounced discordance between physician diagnoses and parental disease perception (P<.001; ). Only 75.6% (68/90) of parents of children with physician-confirmed asthma acknowledged the diagnosis, while merely 57.1% (32/56) of those whose children were ruled out for asthma accepted this conclusion. This difference in diagnostic consistency was most pronounced in the suspected asthma group (discordance rate: 76/134, 56.7%), highlighting a major communication gap in clinical management.

Beyond this discordance, analysis of clinical and environmental factors revealed several patterns more common in the confirmed and suspected asthma groups. Respiratory infections were the predominant wheezing trigger in these groups (62/90, 68.9% and 84/134, 62.7%, respectively). Similarly, seasonal exacerbations in spring and winter were notably more frequent. Inhaled medication use was reported in 60.3% (135/224) of these combined groups, compared to 52.5% (147/280) in the entire cohort. Despite this, 22.3% of patients had uncontrolled symptoms (ACT or C-ACT < 20), indicating a substantial need for improved treatment adherence and effectiveness in real-world practice.

Determinants of Diagnostic Outcomes

Three multivariable logistic regression models were constructed to identify factors associated with asthma diagnostic categories in this cross-sectional cohort (A). In Model 1 (suspected vs excluded asthma), respiratory infection as a wheezing trigger (odds ratio [OR] 4.41, 95% CI 2.16-9.42, P<.001), family history of allergic rhinitis (OR 2.27, 95% CI 1.08-4.99, P=.03), and higher blood eosinophil counts (per 100 cells/μL, OR 1.32, 95% CI 1.05-1.73, P=.02) were significantly associated with suspected asthma (B). In contrast, bronchitis was associated with reduced odds of suspected asthma (OR 0.30, 95% CI 0.14-0.65, P=.002).

‎

Model 2 (confirmed vs suspected asthma) showed that confirmed asthma was associated with older age (per year, OR 1.29, 95% CI 1.14-1.47, P<.001), allergic rhinitis (OR 4.06, 95% CI 1.99-8.31, P<.001), and aeroallergen sensitization (OR 3.83, 95% CI 1.91-7.66, P<.001). Bronchitis remained negatively associated with confirmed asthma (OR 0.21, 95% CI 0.06-0.80, P=.02). These association patterns were consistent in Model 3 (confirmed vs excluded asthma). The robustness of these findings was confirmed through sensitivity analyses (), supporting the clinical validity of the diagnostic stratification and highlighting specific features that distinguish asthma from alternative diagnoses.

Discussion

Principal Findings

This study describes the implementation of a prospective, longitudinal, and digitally enhanced real-world cohort framework that incorporated symptom-driven recruitment, standardized EMRs, ePROs, and residual biospecimen collection within a routine outpatient setting. Leveraging this framework, we enrolled a diagnostically heterogeneous population of children, with a focus on those younger than 6 years, and characterized their distinct clinical profiles according to categories of confirmed, suspected, or excluded asthma. These findings elucidate the broad spectrum of disease presentation in early life and highlight subgroups that have been historically challenging to systematically study, particularly in outpatient settings.

Characteristics of Early-Life Asthma

The current data from the CRCA study provide an initial insight into the characteristics of young children presenting with asthmatic symptoms in an outpatient clinic. The high proportion of participants classified as suspected asthma (179/396, 45.2%), who were also the youngest at enrollment, reflects the diagnostic uncertainty common in early life. This finding aligns with a previous study using the term “reactive airway disease,” which reported a similarly high prevalence of 61.8% (249/403) in a comparable population []. The clinical profile of this group, which showed a lower prevalence of coexisting conditions such as allergic rhinitis and a lower rate of aeroallergen sensitization compared to children with confirmed asthma, yet also differed from the excluded asthma group in features such as lower rates of infectious conditions, suggesting that they may represent a distinct clinical group. This is particularly relevant because although the GINA 2025 criteria provide a definition for suspected asthma in young children, diagnosis remains primarily based on clinical history and lacks reliable biomarkers to support stratification []. Thus, the CRCA framework, with its integrated biospecimen repository, is well-positioned to offer future insights in this area. This observation warrants further longitudinal investigation to understand its implications for disease progression and management.

A notable finding was the discrepancy between physician diagnoses and parental perception of the child’s condition. The fact that a substantial proportion of parents did not concur with the physician’s diagnosis, regardless of whether asthma was confirmed or excluded, points to potential challenges in clinical communication and family education [,]. This discordance was most pronounced in the suspected asthma group (76/134, 56.7%), which is highly important since misalignment in perception in this particular group may lead to both under- and overtreatment, affecting disease management and medication use. This disconnect may be one factor influencing treatment adherence and outcomes in this age group.

A further focus of the CRCA is on early-life environmental exposures [], which are currently captured largely through ePROs. Although no statistically significant differences were observed in the current group comparisons, this is likely due to insufficient sample size. Notably, the suspected asthma group showed the highest reported rate of household water leakage (29/134, 21.6%, P=.11) as well as the highest frequency of humidifier use (33/134, 24.6%, P=.13). These findings are of potentially important given that residential environments in Southwest China are already relatively humid, conditions that are closely associated with mold sensitization and asthma development in children [-]. Furthermore, CRCA’s emphasis on investigating household and personal disinfection and hygiene behaviors aligns with the theoretical frameworks of the hygiene hypothesis and the epithelial barrier hypothesis. As the sample size increases, this focus may support more detailed subgroup analyses in the future. In future phases, in addition to blood specimens, the CRCA plans to include the collection and biobanking of indoor dust samples to further enrich the repository. We also intend to expand our data collection to include maternal health and perinatal factors, thereby providing a more comprehensive assessment of early-life exposures.

To better characterize the distinguishing features of children across diagnostic categories, we constructed three regression models to simulate clinical diagnostic pathways. It must be noted, however, that these models were based on cross-sectional data and thus do not imply causality. We found that infectious triggers of wheezing, a family history of allergic rhinitis, and peripheral blood eosinophil count were the factors most supportive of a suspected asthma diagnosis; these three factors are not currently emphasized in the GINA 2025 guidelines []. Moving further along the diagnostic spectrum, older age, comorbid allergic rhinitis, and aeroallergen sensitization were identified as the factors most strongly associated with a confirmed asthma diagnosis. Furthermore, a coexistence of bronchitis was the primary factor arguing against asthma, which is consistent with GINA criteria. These findings align broadly with current clinical understanding [], yet underscore the particular importance of comorbid allergic rhinitis and sensitization. Moreover, older age emerged as a strongly supportive factor for confirmed asthma, likely reflecting the gradual development of lung function and allergen sensitization throughout childhood. Although this may also be influenced by current diagnostic processes, it reinforces the notion that asthma diagnosis should be viewed through a developmental lens [], potentially requiring a period of ongoing assessment rather than being considered a binary outcome determined in a single visit.

Methodological Considerations Regarding the CRCA Framework

The CRCA framework demonstrates a practical operational model tailored to high-demand clinical environments like China’s, where implementing research poses substantial challenges []. Its shared-responsibility approach, which distributes data and biospecimen collection tasks between clinicians and participants, helps sustain operations within high-volume outpatient workflows. Rather than merely listing its components, the methodological value of CRCA lies in how these features collectively address specific gaps in real-world childhood asthma research. This is evident when contrasted with established international registries (), which often prioritize confirmed diagnoses and may lack integrated biospecimen collection, but have substantially larger sample sizes than the CRCA [-].

The symptom-driven recruitment strategy, which is seldom used in conventional registries that typically require confirmed diagnoses, enables the enrollment of a heterogeneous patient population. This approach mirrors the diagnostic uncertainty common in early-life asthma and allows CRCA to capture disease features from the earliest symptomatic stages. Standard EMR systems are designed for clinical rather than research use, which often results in fragmented and incomplete data [,]. To address this inherent limitation, CRCA implemented standardized EMR templates and a preconsultation data collection step. This approach helps bridge the gap between clinical efficiency and research data quality in time-constrained environments. Furthermore, while ePROs are often optional or partially implemented in other registries, they constitute a core component of CRCA, which enhances its potential scalability and enables richer patient-centric data capture [,]. Similarly, biospecimen collection remains uncommon in large-scale registries due to clinical and logistical burdens; accordingly, CRCA’s residual sample-based model offers a pragmatic and scientifically valuable alternative [-,]. This composition of interoperable, standardized modules suggests that portability to other clinical settings is possible.

It is important to position CRCA appropriately among existing studies. While large-scale, resource-intensive cohorts such as COPSAC have generated landmark findings through long-term follow-up and substantial infrastructure [], CRCA operates within a different context and at a different developmental stage. Its value lies not in direct comparison, but in its seamless integration into routine clinical practice. This integration supports dynamic enrollment and scalable, real-world implementation, aligning with contemporary recommendations for integrating multiomics platforms and target trial methodologies to advance precision medicine under pragmatic conditions []. With further refinement, this clinic-embedded model may offer an adaptable framework for real-world evidence generation, especially in resource-constrained settings [,].

A challenge observed in this study was the initial follow-up rate (48/106, 45.3% among confirmed cases and 28/138, 20.3% among suspected cases), which reflects broader difficulties in pediatric asthma research. Contributing factors included the geographical dispersion of patients, gaps in parental awareness, and the inherent constraints of a regional referral center operating within a tiered health care system. These issues are not unique to CRCA, as comparable studies in the United States have reported similar follow-up rates (22.6%-55.0%) [-]. These realities underscore the importance of developing digital solutions for remote monitoring and self-management—a key future direction for CRCA [,]. In response, we have initiated the development of a digital medicine platform within the established CRCA WeChat public account. This platform is designed to integrate educational content, a mini-program for patient reporting, and an AI agent for interactive questions and answers, aiming to improve follow-up adherence and long-term engagement []. Additionally, the identification of an “off-day follow-up pattern” also suggests a need for more flexible scheduling options, such as night clinics, to improve adherence to review visits [,].

Looking forward, the “digitally enhanced” framework of CRCA could be positioned to support future innovation. We envision that the integrated data from EMRs and ePROs may serve as a basis for developing AI models that could aid in risk stratification. Similarly, the WeChat platform can be leveraged to extend care beyond the clinic, for instance, through remote monitoring, educational content, and AI-facilitated questions and answers. As these digital tools mature, CRCA could evolve into a testbed for evaluating scalable solutions that may improve asthma care in diverse clinical settings.

Limitations

Several limitations should be considered when interpreting our findings. First, as a single-center study conducted in Southwest China, the CRCA cohort primarily comprises patients from Chongqing and neighboring provinces, which may limit the generalizability of our findings to other geographic or socioeconomic contexts. Second, the symptom-driven, outpatient-based recruitment strategy may introduce selection bias, as enrolled children likely represent a more symptomatic or health-seeking population than the general community. Third, the requirement for blood testing (serum sIgE or other tests), though aligned with the goal of residual biospecimen banking, may have excluded children with already confirmed asthma who were under routine treatment, further contributing to selection bias. Fourth, although the CRCA model successfully addressed recruitment challenges, follow-up rates remained suboptimal, particularly among younger children and those with uncertain diagnoses. This attrition may bias longitudinal analyses if loss to follow-up is associated with disease severity or remission. Finally, while we implemented standardized EMR templates and ePROs to enhance data quality, missingness in certain objective measures (eg, lung function, FeNO) was common and not imputed, which may affect the completeness of phenotypic characterization. This largely reflects age-specific clinical guidelines and efforts to minimize burden; future efforts will therefore prioritize novel, feasible biomarkers to strengthen objective profiling.

Conclusions

The CRCA study demonstrates the feasibility of implementing a prospective, longitudinal, and digitally enhanced real-world cohort within routine pediatric outpatient practice. This initiative successfully captures the diagnostically complex population of children with asthmatic symptoms, with a particular focus on early-life presentations, and delineates a distinct clinical subgroup of children with suspected asthma. By integrating symptom-driven recruitment, standardized digital data collection, and biospecimen banking, the CRCA establishes a scalable platform for real-world evidence generation and may provide a valuable foundation for future research into asthma endotypes and personalized management approaches.

TACOMA, Wash., Oct. 30, 2025 /PRNewswire/ —

Acquisition and Branding Update
Columbia Banking System, Inc. (“Columbia,” the “Company,” “we,” or “our”) closed its acquisition of Pacific Premier Bancorp, Inc. (“Pacific Premier”) on August 31, 2025, elevating Columbia’s deposit market share to a top-10 position in Southern California. The acquisition completes our Western footprint and strengthens our presence as a leading financial institution in the western United States. Our integration efforts are progressing smoothly, and we remain on track to integrate systems in the first quarter of 2026.

Columbia Bank began serving customers under its unified name and brand effective September 1, 2025. The strategic transition streamlines our identity across all business lines, including Columbia Wealth Advisors, Columbia Trust Company, Columbia Private Bank, and Columbia Private Trust, making it easier for customers to recognize and engage with the full breadth of our services.

Share Repurchase Authorization Announcement
Columbia’s Board of Directors has authorized the repurchase of up to $700 million of common stock under a new repurchase plan. COLB common share repurchases may be executed in the open market or through privately negotiated transactions, including under Rule 10b5-1 plans. The timing and exact amount of common share repurchases will be at the discretion of senior management and subject to various factors, including, without limitation, Columbia’s capital position, financial performance, market conditions, and regulatory considerations. The repurchase program does not obligate Columbia to purchase any particular number of shares. The authorization will expire on November 30, 2026, but may be suspended, terminated or modified by the Board at any time.

“Our excess capital position as of September 30, 2025 supports the return of additional capital to our shareholders through share repurchases,” commented Mr. Stein. “In addition, we expect to produce exceptional profitability, which will result in meaningful capital generation over the coming quarters. Even as we expand our capital return platform, we are continuing to drive organic growth as we optimize the balance sheet, in line with our commitment to enhancing long-term shareholder value.”

Net Interest Income
Net interest income was $505 million for the third quarter of 2025, up $59 million from the prior quarter. The increase largely reflects the impact of one month operating as a combined company in the current period. Lower interest expense due to a favorable shift in Columbia’s funding mix also contributed to the increase.

Columbia’s net interest margin was 3.84% for the third quarter of 2025, up 9 basis points from the second quarter of 2025. Net interest margin benefited from lower funding costs, due to an increase in customer deposits and corresponding reduction in higher-cost funding sources. The net interest margin was also impacted by one month operating as a combined company in the current period.

The cost of interest-bearing deposits decreased 9 basis points from the prior quarter to 2.43% for the third quarter of 2025, compared to 2.29% for the month of September and 2.20% as of September 30, 2025, reflecting our proactive management of deposit rates ahead of and following the 25-basis point reduction in the federal funds rate in mid-September and a reduction in higher-cost brokered deposits during the month. The cost of interest-bearing deposits in September also benefited from the amortization of a premium related to Pacific Premier’s time deposits, which will continue through December 31, 2025 at an equivalent monthly amount. The amortization contributed $4 million to net interest income during September, and favorably impacted deposit rates. Excluding this impact, the cost of interest-bearing deposits was 2.41% for the month of September and 2.32% as of September 30, 2025.

Columbia’s cost of interest-bearing liabilities decreased 13 basis points from the prior quarter to 2.65% for the third quarter of 2025, compared to 2.47% for the month of September and 2.39% as of September 30, 2025. Excluding the previously discussed premium amortization, the cost of interest-bearing liabilities was 2.58% for the month of September and 2.50% as of September 30, 2025. We expect the premium to be fully amortized by December 31, 2025. Please refer to the Q3 2025 Earnings Presentation for additional net interest margin change details and interest rate sensitivity information.

Non-interest Income
Non-interest income was $77 million for the third quarter of 2025, up $12 million from the prior quarter. The increase was driven by quarterly changes in fair value adjustments and mortgage servicing rights (“MSR”) hedging activity, due to interest rate fluctuations during the quarter, collectively resulting in a net fair value gain of $5 million in the third quarter compared to a net fair value loss of $1 million in the second quarter, as detailed in our non-GAAP disclosures. Excluding these items, non-interest income was up $6 million² between periods, due to one month operating as a combined company.

Non-interest Expense
Non-interest expense was $393 million for the third quarter of 2025, up $115 million from the prior quarter, due to higher merger expense and one month operating as a combined company. Excluding merger and restructuring expense and a $1 million reversal of prior FDIC assessment expense, non-interest expense was $307 million², up $37 million from the prior quarter, as Pacific Premier contributed $34 million to the quarter’s run rate. Other miscellaneous expenses also trended higher as we reinvest prior cost savings into our franchise. Please refer to the Q3 2025 Earnings Presentation for additional expense details.

Balance Sheet
Total consolidated assets were $67.5 billion as of September 30, 2025, up from $51.9 billion as of June 30, 2025, due to the addition of Pacific Premier, partially offset by balance sheet optimization activity in the quarter. Cash and cash equivalents were $2.3 billion as of September 30, 2025, up from $1.9 billion as of June 30, 2025. Including secured off-balance sheet lines of credit, total available liquidity was $26.7 billion as of September 30, 2025, representing 40% of total assets, 48% of total deposits, and 130% of uninsured deposits. Available-for-sale securities, which are held on balance sheet at fair value, were $11.0 billion as of September 30, 2025, an increase of $2.4 billion relative to June 30, 2025, as securities acquired from Pacific Premier and an increase in the fair value of the portfolio was partially offset by net sales during the quarter. Please refer to the Q3 2025 Earnings Presentation for additional details related to our securities portfolio and liquidity position.

Gross loans and leases were $48.5 billion as of September 30, 2025, an increase of $10.8 billion relative to June 30, 2025, due to the addition of Pacific Premier, partially offset by run-off in commercial development and transactional loans, as well as the transfer of $282 million in residential real estate loans to the held-for-sale portfolio.  Excluding these factors, the loan portfolio was essentially unchanged between June 30, 2025 and September 30, 2025. “Our teams continue to focus on new client acquisition and relationship-building, contributing to the 19% increase in new loan originations for the current quarter compared to the prior quarter,” commented Chris Merrywell, President of Columbia Bank. “We continue to prioritize balance sheet optimization and profitability, as we reduce our exposure to non-relationship loans.” Please refer to the Q3 2025 Earnings Presentation for additional details related to our loan portfolio, which include underwriting characteristics, the composition of our commercial portfolios, and disclosure related to transactional loans.

Total deposits were $55.8 billion as of September 30, 2025, an increase of $14.0 billion relative to June 30, 2025, due to the addition of Pacific Premier and organic growth in customer deposits, partially offset by lower brokered deposits. “Customer deposit growth approached $800 million organically during the quarter, reflecting new customer activity and a seasonal lift in balances,” stated Mr. Merrywell. “Our focus on relationship banking directly contributed to new deposit generation in the quarter, which reduced our reliance on wholesale funding sources.” Brokered deposits and borrowings were $4.8 billion as of September 30, 2025, a decrease of $1.9 billion relative to June 30, 2025. Please refer to the Q3 2025 Earnings Presentation for additional details related to deposit characteristics and flows.

Credit Quality
The allowance for credit losses (“ACL”) was $492 million, or 1.01% of loans and leases, as of September 30, 2025, compared to $439 million, or 1.17% as of June 30, 2025. The $53 million increase in the ACL includes the addition of $5 million related to Pacific Premier purchased credit deteriorated (“PCD”) loans, which was booked at acquisition closing and did not affect the income statement. The provision for credit losses was $70 million for the third quarter of 2025 and includes an initial provision for acquired non-PCD loans and unfunded commitments and a recalibration of our models to incorporate historical Pacific Premier data into our ACL assumptions, where applicable. Excluding these items, our provision expense was $0 for the third quarter of 2025².

Net charge-offs were 0.22% of average loans and leases (annualized) for the third quarter of 2025, compared to 0.31% for the second quarter of 2025. Net charge-offs in the FinPac portfolio were $16 million in the third quarter, compared to $14 million in the second quarter. Net charge-offs excluding the FinPac portfolio were $6 million in the third quarter, compared to $15 million in the second quarter. Non-performing assets were $199 million, or 0.29% of total assets, as of September 30, 2025, compared to $180 million, or 0.35% of total assets, as of June 30, 2025. Please refer to the Q3 2025 Earnings Presentation for additional details related to the allowance for credit losses and other credit trends.

Capital
Columbia’s book value per common share was $26.04 as of September 30, 2025, compared to $25.41 as of June 30, 2025. The increase reflects common shares issued and exchanged as a result of the acquisition, net capital generation from operations, and a favorable change in accumulated other comprehensive (loss) income (“AOCI”) to $(268) million as of September 30, 2025, compared to $(333) million as of the prior quarter-end. The change in AOCI is due primarily to a decrease in the tax-effected net unrealized loss on available-for-sale securities to $240 million as of September 30, 2025, compared to $311 million as of June 30, 2025. Tangible book value per common share³ was $18.57 as of September 30, 2025, compared to $18.47 as of June 30, 2025. The items discussed above offset 1.7% tangible book value dilution as a result of the Pacific Premier acquisition, resulting in net tangible book value expansion during the quarter.

Columbia’s estimated total risk-based capital ratio was 13.4% and its estimated common equity tier 1 risk-based capital ratio was 11.6% as of September 30, 2025, compared to 13.0% and 10.8%, respectively, as of June 30, 2025. Columbia remains above current “well-capitalized” regulatory minimums. The regulatory capital ratios as of September 30, 2025 are estimates, pending completion and filing of Columbia’s regulatory reports. 

Earnings Presentation and Conference Call Information
Columbia’s Q3 2025 Earnings Presentation provides additional disclosure. A copy will be available on our investor relations page: www.columbiabankingsystem.com.

Columbia will host its third quarter 2025 earnings conference call on October 30, 2025 at 2:00 p.m. PT (5:00 p.m. ET). During the call, Columbia’s management will provide an update on recent activities and discuss its third quarter 2025 financial results. Participants may join the audiocast or register for the call using the link below to receive dial-in details and their own unique PINs. It is recommended you join 10 minutes prior to the start time.

Join the audiocast: https://edge.media-server.com/mmc/p/i6z93t5w/
Register for the call: https://register-conf.media-server.com/register/BIde1295f868b04a969240d44867cade1a
Access the replay through Columbia’s investor relations page: https://www.columbiabankingsystem.com/news-market-data/event-calendar/default.aspx

About Columbia Banking System, Inc.
Columbia Banking System, Inc. (Nasdaq: COLB) is headquartered in Tacoma, Washington and is the parent company of Columbia Bank, an award-winning western U.S. regional bank. Columbia Bank is the largest bank headquartered in the Northwest and one of the largest banks headquartered in the West with locations in Arizona, California, Colorado, Idaho, Nevada, Oregon, Utah, and Washington. Columbia Bank combines the resources, sophistication, and expertise of a national bank with a commitment to deliver superior, personalized service. The bank supports consumers and businesses through a full suite of services, including retail and commercial banking, Small Business Administration lending, institutional and corporate banking, and equipment leasing. Columbia Bank customers also have access to comprehensive investment and wealth management expertise as well as healthcare and private banking through Columbia Wealth Management. Learn more at www.columbiabankingsystem.com.

Forward-Looking Statements
This press release includes forward-looking statements within the meaning of the “Safe-Harbor” provisions of the Private Securities Litigation Reform Act of 1995, which management believes are a benefit to shareholders. These statements are necessarily subject to risk and uncertainty and actual results could differ materially due to various risk factors, including those set forth from time to time in our filings with the Securities and Exchange Commission. You should not place undue reliance on forward-looking statements and we undertake no obligation to update any such statements. Forward-looking statements can be identified by words such as “anticipates,” “intends,” “plans,” “seeks,” “believes,” “estimates,” “expects,” “target,” “projects,” “outlook,” “forecast,” “will,” “may,” “could,” “should,” “can” and similar references to future periods. In this press release we make forward-looking statements about strategic and growth initiatives and the result of such activity. Risks and uncertainties that could cause results to differ from forward-looking statements we make include, without limitation: current and future economic and market conditions, including the effects of declines in housing and commercial real estate prices, high unemployment rates, continued or renewed inflation and any recession or slowdown in economic growth particularly in the western United States; economic forecast variables that are either materially worse or better than end of quarter projections and deterioration in the economy that could result in increased loan and lease losses, especially those risks associated with concentrations in real estate related loans; risks related to our acquisition of Pacific Premier (the “Transaction”), including, among others, (i) diversion of management’s attention from ongoing business operations and opportunities, (ii) cost savings and any revenue or expense synergies from the Transaction may not be fully realized or may take longer than anticipated to be realized, (iii) deposit attrition, customer or employee loss, and/or revenue loss as a result of the Transaction, and (iv) shareholder litigation that could negatively impact our business and operations; the impact of proposed or imposed tariffs by the U.S. government and retaliatory tariffs proposed or imposed by U.S. trading partners that could have an adverse impact on customers; our ability to effectively manage problem credits; the impact of bank failures or adverse developments at other banks on general investor sentiment regarding the liquidity and stability of banks; changes in interest rates that could significantly reduce net interest income and negatively affect asset yields and valuations and funding sources; changes in the scope and cost of FDIC insurance and other coverage; our ability to successfully implement efficiency and operational excellence initiatives; our ability to successfully develop and market new products and technology; changes in laws or regulations; potential adverse reactions or changes to business or employee relationships; the effect of geopolitical instability, including wars, conflicts and terrorist attacks; and natural disasters and other similar unexpected events outside of our control. We also caution that the amount and timing of any future common stock dividends or repurchases will depend on the earnings, cash requirements and financial condition of Columbia, market conditions, capital requirements, applicable law and regulations (including federal securities laws and federal banking regulations), and other factors deemed relevant by Columbia’s Board of Directors, and may be subject to regulatory approval or conditions.

Non-GAAP Financial Measures
In addition to results presented in accordance with generally accepted accounting principles in the United States of America (“GAAP”), this press release contains certain non-GAAP financial measures. The Company believes presenting certain non-GAAP financial measures provides investors with information useful in understanding our financial performance, our performance trends, and our financial position. We utilize these measures for internal planning and forecasting purposes, and operating pre-provision net revenue and operating return on tangible common equity are also used as part of our incentive compensation program for our executive officers. We, as well as securities analysts, investors, and other interested parties, also use these measures to compare peer company operating performance. We believe that our presentation and discussion, together with the accompanying reconciliations, provides a complete understanding of factors and trends affecting our business and allows investors to view performance in a manner similar to management. These non-GAAP measures should not be considered a substitution for GAAP basis measures and results, and we strongly encourage investors to review our consolidated financial statements in their entirety and not to rely on any single financial measure. Because non-GAAP financial measures are not standardized, it may not be possible to compare these financial measures with other companies’ non-GAAP financial measures having the same or similar names.

SOURCE Columbia Banking System, Inc.