Disparities in Cough Variant Asthma Management and Physician-Reported Pulmonary Function Testing Resources: A Cross-Sectional Survey of Respiratory Physicians Across Hospital Tiers in Zhejiang, China Authors: Jiangying Guo, Xi Zhang, Jianxing Lai, Jian Wang, Linfeng Shen, Yonghong Zhong, Huaqiong Huang Published: 2026 ABSTRACT AbstractPurposeThe study aimed to evaluate differences in cough variant asthma (CVA) management and physician-reported pulmonary function testing resources across hospital tiers in Zhejiang, China.MethodsFrom October 12 to November 15, 2025, a cross-sectional online questionnaire survey was conducted among physicians engaged in respiratory or chronic cough care in Zhejiang, China. The survey assessed physicians’ knowledge of CVA diagnosis and treatment, awareness of prognosis, and reported access to pulmonary function testing resources. Responses were compared across hospital tiers.ResultsA total of 550 submitted questionnaires were received, and 522 valid responses were included after data cleaning. Physicians in tertiary hospitals were most accurate in diagnosing CVA according to guideline criteria (85.79%), compared with those in secondary (70.59%) and primary hospitals (51.53%). Reported access to key tests, especially bronchial provocation testing, was lower in secondary and primary hospitals. The proportion of physicians who intended to use bronchial provocation testing exceeded the proportion who actually performed it (41.57% vs 23.75%). For initial treatment, 91.05% of tertiary hospital physicians selected inhaled corticosteroids (ICS) combined with long-acting beta2 agonists (LABA). This was significantly higher than the proportions in secondary and primary hospitals (69.12% and 37.24%, respectively). Physicians in tertiary hospitals also showed greater awareness of CVA prognosis.ConclusionClear differences exist in CVA care across hospital tiers in Zhejiang, China. Lower-tier hospitals showed gaps in diagnosis, access to bronchial provocation testing, and ICS/LABA selection. In addition, the gap between intended and performed use of bronchial provocation testing may reflect a potential barrier to objective CVA diagnosis.Keywords: cough variant asthma, chronic cough, pulmonary function test, hospital tiers, guideline adherence, clinical practice Introduction Chronic cough is characterized by continuous coughing that lasts for over eight weeks in adults, representing a significant global public health concern that affects approximately 10% of the adult population worldwide.1 As one of the most common chief complaints in respiratory outpatient clinics, chronic cough has a prevalence of approximately 6.2% among Chinese adults, accounting for 10%–38% of respiratory outpatient visits.2,3 Cough variant asthma (CVA) is one of the leading causes of chronic cough, accounting for approximately one-third of chronic cough etiologies. It features “paroxysmal cough as the sole or main symptom, without typical asthmatic manifestations such as wheezing or chest tightness.”4,5 The diagnosis and treatment of CVA are highly dependent on clinical guidelines. Both Chinese and international guidelines clearly stipulate that CVA diagnosis meets four core criteria: “symptom assessment + evidence of reversible airflow limitation + responsiveness to asthma treatment + exclusion of other cough etiologies”.4,6–10 Among these, pulmonary function tests (eg, spirometry, bronchial provocation and bronchodilation tests) are critical for confirming reversible airflow limitation and airway hyperresponsiveness. Compared with classic asthma, atypical asthma such as CVA is often characterized by relatively preserved baseline lung function.11–15 Bronchodilation tests therefore have a low positive rate,13 and a negative result cannot exclude CVA. In this context, bronchial provocation testing remains important for objective confirmation of CVA by demonstrating airway hyperresponsiveness.16,17 When such objective evidence is lacking, CVA can be easily confused with common chronic cough etiologies such as gastroesophageal reflux-induced cough, upper airway cough syndrome, and atopic cough, leading to misdiagnosis or underdiagnosis.5,18 Additionally, guidelines recommend inhaled corticosteroids (ICS) combined with long-acting beta2 agonists (LABA) as the first-line treatment for CVA, with a treatment duration of more than 8 weeks. Approximately 30%–40% of adult patients with untreated CVA may develop classic asthma within several years.19 Early ICS treatment may reduce this risk and prevent the subsequent development of typical symptoms, such as wheezing and dyspnea.20,21 However, the therapeutic effect in CVA patients also relies on physicians’ scientific assessment of efficacy (eg, asthma control test [ACT] scores and pulmonary function test [PFT]) and attention to comorbidities.4,9,10 In recent years, China has been gradually promoting the standardized diagnosis and treatment of respiratory diseases at the primary care level, but significant hierarchical heterogeneity exists in clinical practice across hospitals of different levels.22,23 The China Pulmonary Health Study reported that, among an estimated 45.7 million adults with asthma in China, only 23.4% had ever undergone lung function testing and only 28.8% had been diagnosed by physicians.24 Similar tier-related gaps have also been reported in chronic airway disease management, with the rate of PFT use at initial diagnosis decreasing from 99.0% in tertiary teaching hospitals to 69.4% in secondary hospitals and 29.9% in primary healthcare institutions.25 These findings suggest that the uneven distribution of medical resources, limited use of pulmonary function testing, and lack of equipment in primary medical institutions may lead to insufficient objective evidence for CVA diagnosis.22,26,27 In addition to diagnostic resources, treatment resources may also be unevenly distributed. A nationwide survey of 1018 Chinese hospitals showed that most inhaled drugs had an availability rate below 20% in primary hospitals.28 Moreover, physicians in hospitals of different levels may have discrepancies in their awareness of CVA diagnostic criteria and treatment practices, which may further affect the effectiveness of CVA management.29–31 A previous nationwide survey in China reported differences in atypical asthma management between specialists and primary care practitioners.32 However, no prior study has systematically stratified CVA diagnostic knowledge, treatment practices, and pulmonary function resources by hospital tier in China. Because CVA usually presents as chronic cough without typical asthma symptoms, patients may first attend chronic cough or general respiratory clinics, increasing the risk of under-recognition or delayed diagnosis. Therefore, we surveyed 550 physicians from hospitals at different levels in Zhejiang province, China, aiming to: (1) assess physicians’ cognition of CVA diagnostic criteria and treatment practices across hospital tiers to identify discrepancies from current guidelines; (2) investigate the status and deficiencies of PFT equipment configuration in hospitals of different levels; (3) explore the association between PFT accessibility and standardized CVA management to provide evidence for optimizing the tiered healthcare system. Material and Methods Study Design We conducted a cross-sectional online questionnaire survey in Zhejiang Province, China, from October 12 to November 15, 2025, using convenience sampling. The questionnaire was distributed via WeChat by regional respiratory expert members of the Zhejiang Provincial Cough Alliance to physicians working in local hospitals of different tiers. The Zhejiang Provincial Cough Alliance is a provincial medical collaborative platform composed of respiratory specialists from all 11 prefecture-level cities in Zhejiang Province, aiming to promote standardized chronic cough diagnosis and clinical management. To reduce potential overrepresentation of individual hospitals, regional expert members were instructed to forward the questionnaire to no more than two eligible physicians from the same hospital whenever possible. Participants The participants were a convenience sample of physicians engaged in respiratory or chronic cough care from hospitals of different tiers in Zhejiang Province. Hospital tiers were classified according to the Chinese hospital grading system. Primary hospitals generally provide basic medical, preventive, and rehabilitative services and usually have fewer than 100 beds. Secondary hospitals are regional hospitals that provide comprehensive medical and health services, usually with 100–499 beds. Tertiary hospitals are major referral centers at the city, provincial, or national level, usually with at least 500 beds, and provide high-level specialist care as well as medical education and research. A total of 550 submitted questionnaires were received, including 198 from tertiary hospitals, 141 from secondary hospitals, and 211 from primary hospitals. Because the total number of questionnaire visits could not be determined, the exact response rate could not be calculated. Two data-cleaning procedures were applied to ensure data quality and response reliability. First, two similar questions regarding bronchial provocation test availability were incorporated in distinct sections of the questionnaire to evaluate response consistency; questionnaires with inconsistent answers to these repeated items were excluded. Second, based on the estimated minimum time required to complete the survey thoughtfully, responses completed and submitted in <90 seconds were deemed unreliable and excluded from the final analysis. Following these data-cleaning procedures, 12 questionnaires were excluded for inconsistent responses to repeated items, and 16 for completion times under 90 seconds. A total of 522 valid questionnaires were finally included for analysis, with a valid questionnaire rate of 94.9%. Participation was voluntary and uncompensated. To protect participant and institutional privacy, the questionnaire was anonymous and did not collect physicians’ names, hospital names, or hospital identifiers. Therefore, all analyses were based on individual physician responses. PFT availability and equipment data were based on physician self-reports. Questionnaire Design and CVA Diagnostic Criteria No validated questionnaire specifically suitable for this study was available. Therefore, two chief physicians in respiratory medicine from the research team developed the initial draft based on current Chinese guidelines for bronchial asthma and chronic cough. The draft was subsequently reviewed and revised by respiratory experts from the Zhejiang Provincial Cough Alliance. Each item was evaluated by the expert panel as essential, useful, or irrelevant to the study objectives. After all comments and suggestions were incorporated, the questionnaire was finalized for distribution. The questionnaire covered physicians’ knowledge of CVA diagnostic criteria, diagnostic and treatment practices, prognosis awareness, and self-reported access to pulmonary function testing resources. The full questionnaire is provided in Figure S1. Online surveys were distributed and collected via WeChat. The diagnostic criteria for CVA were based on the Guidelines for the Prevention and Management of Bronchial Asthma (2024 Edition).4 Specifically, these criteria include: (1) cough is the only or main symptom, without wheezing, shortness of breath, and other symptoms and signs of typical asthma; (2) examinations to support reversible airflow limitation; (3) except for other diseases that cause cough; (4) it is effective to treat the disease according to the treatment for typical asthma. Statistical Analysis All variables in the questionnaire were categorical and summarized as frequencies with percentages. For comparisons among the three hospital tiers (tertiary, secondary, and primary), the Kruskal–Wallis test was used for ordinal variables, and the Chi-squared test (or Fisher’s exact test when expected cell counts were <5) was applied for nominal categorical variables. For overall comparisons, a two-tailed P < 0.05 was considered statistically significant. If the overall comparison was statistically significant, pairwise post-hoc comparisons were conducted using the Mann–Whitney U-test for ordinal data and Chi-squared or Fisher’s exact tests for nominal categorical data. All pairwise comparisons were adjusted for multiple comparisons using the Bonferroni correction, with the adjusted significance level set at α = 0.05/3 = 0.017. Multivariable logistic regression analyses were performed to assess the associations between hospital tier and core diagnostic and treatment strategies for CVA. Tertiary hospitals were used as the reference category. The models were adjusted for professional title, sex, years of service, educational background, pulmonary and critical care medicine (PCCM) department, cough subspecialty program, and asthma clinic. Odds ratios (ORs) with 95% confidence intervals (CIs) and P-values were reported. For multivariable logistic regression analyses, a two-tailed P < 0.05 was considered statistically significant. All statistical analyses were performed using SPSS 26.0 software (IBM Corp., Armonk, NY, USA). Figures were further plotted in GraphPad Prism 10.0 (GraphPad Software, San Diego, CA, USA) for visualization. Results Characteristics of Survey Participants and Hospitals A total of 522 physicians were included in the final analysis, of whom 190 (36.40%) were from tertiary hospitals, 136 (26.05%) from secondary hospitals, and 196 (37.55%) from primary hospitals. The baseline characteristics of participants and hospitals, stratified by hospital level, are presented in Table 1. Significant differences were observed across hospital tiers in the availability of PCCM departments, cough subspecialty programs, asthma clinics, sex distribution, as well as physicians’ professional titles and educational backgrounds (all P < 0.05, Table 1). No significant difference was observed in years of service among physicians from different hospital tiers. Table 1Baseline Characteristics of Survey Participants and HospitalsVariablesTotal (n = 522)Tertiary Hospital (n = 190)Secondary Hospital (n = 136)Primary Hospital (n = 196)P Value†PCCM department, n (%)250 (47.89)167 (87.89)69 (50.74)14 (7.14)<0.001Subspecialty program in cough, n (%)201 (38.51)120 (63.16)44 (32.35)37 (18.88)<0.001Asthma clinic, n (%)229 (43.87)153 (80.53)56 (41.18)20 (10.20)<0.001Professional title, n (%)<0.001 Chief Physician117 (22.41)64 (33.68)33 (24.26)20 (10.20) Associate Chief Physician134 (25.67)50 (26.32)44 (32.35)40 (20.41) Attending Physician155 (29.69)53 (27.89)42 (30.88)60 (30.61) Resident Physician116 (22.22)23 (12.11)17 (12.50)76 (38.78)Sex, n (%)0.003 Male266 (50.96)92 (48.42)86 (63.24)88 (44.90) Female256 (49.04)98 (51.58)50 (36.76)108 (55.10)Years of service, n (%)0.104 5–10 years167 (31.99)59 (31.05)34 (25.00)74 (37.76) 11–20 years166 (31.80)57 (30.00)51 (37.50)58 (29.59) 21–30 years142 (27.20)58 (30.53)35 (25.74)49 (25.00) 31–40 years39 (7.47)13 (6.84)13 (9.56)13 (6.63) >40 years8 (1.53)3 (1.58)3 (2.21)2 (1.02)Education level, n (%)<0.001 Associate’s degree32 (6.13)1 (0.53)1 (0.74)30 (15.31) Bachelor’s degree366 (70.11)77 (40.53)126 (92.65)163 (83.16) Master’s degree106 (20.31)94 (49.47)9 (6.62)3 (1.53) Doctoral degree18 (3.45)18 (9.47)0 (0.00)0 (0.00)Notes: Data are presented as n (%). † Kruskal–Wallis test was used for ordinal variables; Chi-squared test was used for nominal categorical variables, with Fisher’s exact test used when expected cell counts were <5. Cough subspecialty program, chronic cough subspecialty outpatient services or related clinical program development.Abbreviation: PCCM, pulmonary and critical care medicine. Baseline Characteristics of Survey Participants and Hospitals Notes: Data are presented as n (%). † Kruskal–Wallis test was used for ordinal variables; Chi-squared test was used for nominal categorical variables, with Fisher’s exact test used when expected cell counts were <5. Cough subspecialty program, chronic cough subspecialty outpatient services or related clinical program development. Abbreviation: PCCM, pulmonary and critical care medicine. Disparities in CVA Diagnostic Practices Among chronic cough patients with normal chest imaging, CVA (90.80%) was the most frequently suspected etiology by participating clinicians, followed sequentially by upper airway cough syndrome (UACS) (75.86%) and allergic cough (72.99%). The proportion of all other suspected etiologies was below 70% (Figure S2). Overall, 33.14% and 47.51% of physicians self-reported their knowledge of CVA diagnostic criteria as “very clear” and “mostly clear”. Meanwhile, 19.16% reported it as “slightly clear”, and 0.19% were “unaware”. Significant differences were observed across hospital tiers (P < 0.001, Figure 1A). Pairwise comparisons showed that physicians from tertiary hospitals reported a higher level of understanding compared to those from secondary and primary hospitals (both P < 0.001, Figure 1A). Diagnostic knowledge and practices for CVA among physicians. (A) Physicians’ self-reported knowledge of CVA diagnostic criteria. (B) Accuracy in CVA diagnostic criteria. (C) Proportion of empirical diagnosis among CVA patients as reported by physicians. (D) Proportion of physician-selected diagnostic methods for CVA (%). (E) Reported deficiencies in the diagnosis of CVA. Notes: For overall three-group comparison: P < 0.05 was defined as statistically significant; For pairwise post-hoc comparisons after Bonferroni correction (α=0.05/3=0.017): NS, not significant; *P < 0.017; **P < 0.01; ***P < 0.001. Abbreviations: CVA, cough-variant asthma; PEF, peak expiratory flow; FeNO, fractional exhaled nitric oxide; PFT, pulmonary function test. Question 18 (detailed in Figure S1) aimed to assess physicians’ adherence to the recommended diagnostic criteria for CVA. A physician’s answer was considered correct only if it matched all four parts of the CVA diagnosis checklist from the 2024 Chinese Asthma Guidelines. Overall, 68.97% of physicians chose the correct criteria. However, physicians in tertiary hospitals had a higher accuracy rate (85.79%) than those in secondary hospitals (70.59%) and primary hospitals (51.53%). These differences were statistically significant (both P < 0.001, Figure 1B). We also assessed the frequency of empirical CVA diagnosis when test results were inconclusive. Overall, 43.87% of physicians reported rarely doing this (in fewer than 10% of cases), and 37.36% said they did it occasionally (in 11–30% of cases). Empirical diagnosis frequency did not differ significantly among physicians across hospital types (P = 0.319, Figure 1C). Regarding the selection of diagnostic methods for CVA, 71.46% of physicians utilized medical history inquiry, 70.50% chose pulmonary ventilation function plus bronchodilation tests, 54.98% chose fractional exhaled nitric oxide (FeNO) tests, and only 54.21% selected pulmonary ventilation function plus bronchial provocation tests. The proportion of physicians selecting bronchial provocation testing was significantly higher in tertiary hospitals than in secondary and primary hospitals (both P < 0.001, Figure 1D), whereas no significant difference was observed between secondary and primary hospitals (P = 0.396, Figure 1D). For the deficiencies in CVA diagnosis, insufficient disease awareness was the predominant factor, with no significant difference across the three hospital tiers. This was reported by 85.20% of physicians in tertiary hospitals, 86.03% in secondary hospitals, and 77.89% in primary hospitals. The second was a lack of objective diagnostic tests, noted by 72.96%, 76.47%, and 60.00% of physicians across the three tiers, respectively (Figure 1E). Pulmonary Function Equipment Allocation and Reported Barriers To further investigate the lack of objective diagnostic tests for CVA, we assessed physician-reported PFT equipment availability across hospital tiers. Significant differences were observed in the physician-reported configuration of both portable and laboratory-based PFT devices across hospital levels (both P < 0.001, Figure 2A and B; Table S1). 80.35% of physicians reported that their institutions had at least one portable spirometer. Overall comparisons showed significant differences in the reported number of portable spirometers among the three hospital tiers (P < 0.001, Figure 2A and Table S1). Pairwise comparisons showed that tertiary hospitals had a significantly higher reported number of portable spirometers than secondary hospitals (P = 0.008) and primary hospitals (P < 0.001), whereas no significant difference was observed between secondary and primary hospitals (P = 0.52). For laboratory-based PFT devices, 60.42% of physicians reported that their institutions were equipped with such devices. Specifically, the equipment rate reached 93.14% in tertiary hospitals, 83.06% in secondary hospitals, and only 11.93% in primary hospitals. Pairwise comparisons further revealed significant differences in the quantity of laboratory-based PFT devices between each pair of hospital levels (all P < 0.001, Figure 2B and Table S1). Physician-reported PFT resources and implementation barriers. (A) Reported number of portable PFT devices across hospital tiers. (B) Reported number of laboratory-based PFT devices across hospital tiers. (C) Physician-reported available pulmonary function tests in different hospital tiers. (D) Disparities in intended and performed pulmonary function tests among physicians for initial evaluation. (E) Reported major barriers in PFT implementation. continued. Notes: For overall three-group comparison: P < 0.05 was defined as statistically significant; For pairwise post-hoc comparisons after Bonferroni correction (α=0.05/3=0.017): NS, not significant; *P < 0.017; **P < 0.01; ***P < 0.001. Abbreviations: PFT, pulmonary function test; PEF, peak expiratory flow; FeNO, fractional exhaled nitric oxide. Comparisons were made among the available PFTs at primary, secondary, and tertiary hospitals (Figure 2C). Pulmonary ventilation function was reported to be available by 98.42% of tertiary hospital physicians, 94.12% of secondary hospital physicians, and 77.04% of primary hospital physicians. The bronchodilation test was reported as available by 96.32% of physicians from tertiary hospitals and 85.29% from secondary hospitals, both significantly higher than the 19.90% reported by primary hospital physicians (both P < 0.001). The reported availability of bronchial provocation tests differed significantly across hospital tiers, with 74.21% of tertiary-hospital physicians reporting availability at their institutions, significantly higher than 29.41% of secondary hospital physicians and 3.06% of primary hospital physicians (both P < 0.001). The reported availability of other pulmonary function tests also generally decreased from tertiary to secondary and primary hospitals (Figure 2C). To further assess the gap between intended and performed PFT use in CVA diagnosis, we surveyed physicians on their intended PFTs and those actually ordered for patients (Figure 2D and Table S2). Notably, a significantly higher proportion of physicians intended to perform bronchial provocation testing than actually did so (41.57% vs 23.75%, P < 0.001). No significant discrepancies were found between intended and performed use of bronchodilation tests, FeNO tests, or peak expiratory flow (PEF) variability assessment (Table S2). Stratified analyses further showed that intended and performed PFTs differed significantly within each hospital tier, including tertiary, secondary, and primary hospitals (Figure S3). Regarding barriers to PFT implementation, poor patient willingness was consistently the most frequently reported barrier across all hospital levels (tertiary: 74.21%, secondary: 77.21%, primary: 69.90%; Figure 2E), while the second and third key barriers differed by tiers. For tertiary hospital clinicians, the second and third barriers were self-payment requirement for FeNO tests (71.58%) and lack of clinician attention (46.84%); for secondary hospital clinicians, they were insufficient PFT specialists (54.41%) and lack of clinician attention (50.74%); for primary hospital clinicians, they were insufficient PFT specialists (62.76%) and insufficient PFT equipment (49.49%). The high cost of PFT, poor PFT quality control, and other barriers were also mentioned by physicians (Figure 2E). Disparities in the Initial Treatment of CVA Initial medications for newly diagnosed CVA exhibited differences across hospital levels (Figure 3A). Overall, montelukast sodium was the most frequently selected agent by clinicians (69.16%), followed by ICS/LABA (65.13%) and compound methoxyphenamine (54.60%). Among tertiary hospital clinicians, ICS/LABA was the predominant choice (91.05%), which was significantly higher than that in secondary hospitals (69.12%) and primary hospitals (37.24%) (both P < 0.001). Montelukast sodium was the most common choice in secondary and primary hospitals, selected by 69.85% and 70.92% of physicians, respectively. In tertiary hospitals, 66.84% of physicians selected montelukast sodium, with no significant difference among the three hospital tiers. Initial treatment of CVA by hospital level. (A) Proportion of initial medications for newly diagnosed CVA. (B) Initial ICS-containing regimens for newly diagnosed CVA. Notes: For overall three-group comparison: P < 0.05 was defined as statistically significant; For pairwise post-hoc comparisons after Bonferroni correction (α=0.05/3=0.017): NS, not significant; *P < 0.017; **P < 0.01; ***P < 0.001. Abbreviations: CVA, cough-variant asthma; ICS, inhaled corticosteroid; SABA, short-acting beta2-agonist; LABA, long-acting beta2-agonist; LAMA, long-acting muscarinic antagonist. Initial ICS-containing regimens for newly diagnosed CVA differed significantly across tertiary, secondary, and primary hospital physicians (P < 0.001, Figure 3B). Overall, ICS/LABA was the most commonly selected regimen (58.81%), followed by ICS only (38.89%), while the combination of ICS, LABA, and long-acting muscarinic antagonist (LAMA) was rarely used (2.30%). For ICS/LABA regimens, tertiary hospital physicians had the highest selection rate (72.11%), which was significantly higher than that among primary hospital physicians (42.35%) (P < 0.001, Table S3). Secondary hospital physicians showed an intermediate rate (63.97%), which was also significantly higher than primary hospital physicians (P < 0.001) but not significantly different from tertiary hospital physicians (P = 0.118). Among ICS/LABA subcategories, low-dose ICS/LABA was the most commonly selected across all hospitals. Efficacy Evaluation, Prognostic Awareness, and Perceived Challenges in CVA Management For the selection of efficacy evaluation methods, patients’ subjective complaints were the most commonly selected across all hospital levels (86.21%), followed by cough score (71.65%), PFT (67.8%), and ACT score (50.57%, Figure 4A). Regarding physicians’ self-assessed treatment effectiveness, 49.23% of physicians reported excellent outcomes and 47.32% reported moderate outcomes (Figure 4B). Notably, this self-assessed effectiveness differed significantly across the three hospital levels (P < 0.001). Pairwise comparisons showed that self-assessed effectiveness in tertiary hospitals was significantly higher than that in secondary and primary hospitals (both P < 0.001, Table S4). Efficacy evaluation, treatment outcomes, and awareness of comorbidities and prognosis in CVA. (A) Methods used by physicians to evaluate treatment efficacy in CVA patients. (B) Physicians’ self-assessment of treatment effectiveness for CVA. (C) Reported reasons for poor CVA treatment effect. (D) Physicians’ awareness of CVA comorbidities. (E) Physicians’ self-reported knowledge on CVA prognosis. Notes: For overall three-group comparison: P < 0.05 was defined as statistically significant; For pairwise post-hoc comparisons after Bonferroni correction (α=0.05/3=0.017): NS, not significant; **P < 0.01; ***P < 0.001. Abbreviations: CVA, cough-variant asthma; ACT, asthma control test; VAS, visual analog scale; PFT, pulmonary function test. Overall, the top three reasons for poor treatment effect in CVA were inadequate duration of inhaled medications (76.25%), irregular inhalation technique (76.05%), and repeated allergen exposure (74.71%). Physicians in tertiary hospitals reported irregular inhalation technique (82.63%) as the most common reason, followed by repeated exposure to allergens (80.00%). In secondary hospitals, the top two factors were repeated allergen exposure (81.62%) and inadequate duration of inhaled medications (77.94%). For primary hospitals, the predominant factors were inadequate duration of inhaled medications (72.96%) and irregular inhalation technique (70.41%). Additionally, physicians also mentioned insufficient dosage of inhaled medications, recurrent respiratory infections, and the presence of comorbidities as contributing factors (Figure 4C). The awareness rate of CVA comorbidities was 98.42% in tertiary hospital physicians, 93.38% in secondary hospital physicians, and 88.78% in primary hospital physicians. Overall differences were observed across hospital tiers, mainly driven by the difference between tertiary and primary hospitals (P < 0.001, Figure 4D). Physicians’ self-reported knowledge on CVA prognosis was significantly different among the three hospital types (P < 0.001, Figure 4E). Pairwise comparisons revealed that physicians in tertiary hospitals reported significantly greater knowledge of CVA prognosis than those in secondary hospitals (P = 0.002) and primary hospitals (P < 0.001, Table S4). Multivariable Logistic Regression Analysis of Hospital Tier and Core Diagnostic and Treatment Practices for CVA Given that physician and institutional characteristics differed across hospital tiers, we performed multivariable logistic regression analyses to further assess the associations between hospital tier and core diagnostic and treatment practices for CVA. The models were adjusted for professional title, sex, years of service, education level, PCCM department, cough subspecialty program, and asthma clinic (Table S5). Compared with tertiary hospital physicians, primary hospital physicians were less likely to report guideline-concordant CVA diagnosis (adjusted OR = 0.33, 95% CI: 0.15–0.70, P = 0.004), whereas the difference between secondary and tertiary hospitals was not statistically significant after adjustment (adjusted OR = 0.60, 95% CI: 0.30–1.17, P = 0.134). For the selection of bronchial provocation testing for CVA diagnosis, secondary hospital physicians were less likely than tertiary hospital physicians to select this test (adjusted OR = 0.42, 95% CI: 0.23–0.75, P = 0.003). For initial treatment, both secondary and primary hospital physicians were less likely than tertiary hospital physicians to select ICS/LABA, with adjusted ORs of 0.45 (95% CI: 0.22–0.93, P = 0.032) and 0.19 (95% CI: 0.08–0.42, P < 0.001), respectively. Discussion This research is the first online survey to reveal significant hierarchical disparities in CVA management and PFT resource allocation among different levels of hospitals in Zhejiang, China. These findings could inform similar healthcare systems nationwide. Our study identified significant discrepancies between current clinical guideline recommendations and the clinical practices of primary and secondary hospitals, particularly in diagnostic standardization, PFT selection, and therapeutic medication use. After adjustment for physician and institutional characteristics, hospital tier remained associated with core diagnostic and treatment practices, including guideline-concordant CVA diagnosis and ICS/LABA selection. Although therapeutic practices in tertiary hospitals were largely aligned with guidelines, their diagnostic processes still required further improvement. Our study revealed that CVA was the most frequently considered diagnosis among patients with chronic cough and normal chest imaging. This finding aligned with previous Chinese epidemiological data.4,5 However, significant disparities existed in physicians’ diagnostic awareness and accuracy for CVA across different hospital levels. Physicians in tertiary hospitals demonstrated significantly higher awareness of CVA diagnostic criteria and a higher rate of guideline-consistent diagnosis (85.79%) compared to those in secondary (70.59%) and primary hospitals (51.53%). This discrepancy may be attributed to previous studies suggesting that physicians in primary and secondary hospitals receive less systematic professional training and that there is a lack of evidence and guidelines tailored specifically for primary care.33 Prior studies have shown that specialists exhibit higher guideline adherence than primary care physicians,34–36 which is consistent with the data from our survey. However, the frequency of empirical diagnosis was similar across hospital tiers. This finding suggests that CVA diagnosis remains challenging for physicians at all hospital levels, rather than a problem limited to primary care physicians. Limited access to objective testing may further increase this difficulty in primary and secondary hospitals. Regarding the selection of diagnostic methods, previous studies have established that a positive bronchial provocation test is not only a core tool for confirming airway hyperresponsiveness but also the most important criterion for diagnosing CVA.4,6,37 Our study found that the utilization rate of bronchial provocation tests was significantly higher in tertiary hospitals than in secondary and primary hospitals. Patients with mild or atypical asthma exhibit minimal reductions in forced expiratory volume in 1 second (FEV1) and the ratio of FEV1 to forced vital capacity (FEV1/FVC), which are basically within the normal range.11–15 Consequently, the positive rate of bronchodilation tests in these patients is low, and a negative result cannot exclude asthma.13 Furthermore, a randomized controlled trial indicated that up to one-third of CVA patients exhibit no bronchodilator response.38 This is likely because many CVA patients maintain good pulmonary function without baseline airflow limitation, making it difficult to demonstrate reversibility.39 However, bronchodilation tests and medical history inquiry were the most commonly employed diagnostic methods in primary and secondary hospitals. FeNO testing was also selected by more than half of physicians in our survey. A meta-analysis of 15 studies reported moderate diagnostic accuracy, with a sensitivity of 0.72 and a specificity of 0.85.40 However, FeNO testing cannot replace bronchial provocation testing for confirming airway hyperresponsiveness. Reliance on these methods without bronchial provocation testing may increase the risk of misdiagnosing or underdiagnosing CVA, potentially confusing it with other chronic cough etiologies such as UACS and atopic cough.41,42 Notably, insufficient disease awareness and inadequate access to objective diagnostic tests were reported as the main barriers to CVA diagnosis across all hospital levels. This aligns with previous studies that highlight knowledge gaps and resource limitations as hindrances to standardized asthma diagnosis in primary care settings.32,43 Our study also revealed significant disparities across hospital levels in the availability of pulmonary function tests and in the quantity of pulmonary function equipment (both portable and laboratory-based). Although pulmonary ventilation function tests are widely available in most hospitals, the low availability of bronchial provocation tests (3.06%) and bronchodilation tests (19.90%) in primary hospitals may limit the confirmation of the core diagnostic criterion for CVA. Furthermore, there was a significant discrepancy between physicians’ intended and actual performance of pulmonary function tests in chronic cough patients with normal chest imaging, particularly for bronchial provocation tests (41.57% vs 23.75%). In contrast, no significant gap between intended and performed use was observed for bronchodilation testing, FeNO testing, or PEF variability assessment. Limited access to bronchial provocation testing may therefore be a key barrier to guideline-recommended CVA diagnosis. Resource allocation should prioritize bronchial provocation testing capacity, standardized quality control, and regional referral pathways, rather than general PFT expansion alone. For the initial treatment of CVA, the proportion of physicians selecting antibacterial agents was very low across all hospital levels. However, significant deviations from guideline recommendations were still observed in primary and secondary hospitals. This study found that montelukast sodium was the most commonly selected initial medication for CVA among physicians in secondary (69.85%) and primary hospitals (70.92%). In contrast, physicians in tertiary hospitals (91.05%) primarily chose ICS/LABA. According to Chinese guidelines, ICS/LABA is the first-line treatment for CVA, whereas montelukast sodium is mainly used as adjunctive therapy in selected patients with poor response to ICS or ICS/LABA or with more severe airway inflammation.4,44 Nevertheless, montelukast sodium has been shown to alleviate cough symptoms, improve quality of life, and reduce airway inflammation.45–48 A randomized open-label trial showed that montelukast monotherapy had an antitussive effect similar to that of budesonide/formoterol, although combination treatment provided additional improvement in lung function.48 In our study, montelukast sodium was selected at similar rates across hospital tiers, whereas ICS/LABA selection decreased progressively from tertiary to primary hospitals. This suggests that the main treatment difference across hospital tiers was not whether physicians selected adjunctive medications such as montelukast sodium, but whether they selected ICS/LABA as a core first-line treatment. In addition, compound methoxyphenamine was the third most frequently selected initial medication overall (54.60%). In primary hospitals, it was selected more frequently than ICS/LABA (58.16% vs 37.24%). The Chinese cough guideline recommends compound methoxyphenamine mainly for postinfectious cough, rather than as first-line treatment for CVA.44 This pattern may suggest a greater emphasis on symptomatic treatment than on standard anti-inflammatory therapy in initial CVA management in primary hospitals. Meanwhile, numerous studies have demonstrated that combination treatment alleviates cough symptoms more swiftly and efficiently than either ICS or bronchodilator monotherapy in CVA patients.49–51 For the initial choice of ICS-containing regimens, physicians in tertiary (72.11%) and secondary hospitals (63.97%) primarily selected ICS/LABA, whereas those in primary hospitals (53.57%) mainly chose ICS monotherapy. These disparities may stem from multiple factors, including insufficient awareness of updated guidelines among primary care physicians, the lack of availability of ICS/LABA in primary hospitals, and the higher accessibility of montelukast sodium in these institutions. Questionnaire data showed that treatment effectiveness was most commonly evaluated by patients’ subjective complaints (86.21%) and cough scores (71.65%). In contrast, the Global Asthma Physician Survey reported that only 10% of physicians used validated patient-reported questionnaires to monitor asthma control.52 This difference may partly reflect differences in survey populations and healthcare settings. Our study covered hospitals of different tiers across Zhejiang Province, including a large proportion of primary hospitals, where access to pulmonary function testing and patients’ willingness or ability to pay for repeated objective tests may be more limited. For CVA, this is clinically relevant because symptom improvement alone may not fully reflect airway recovery. Without objective follow-up, such as lung function reassessment, persistent airway dysfunction may be overlooked. Overall, 49.23% and 47.32% of physicians rated CVA treatment effectiveness as excellent and moderate, respectively. Physicians in tertiary hospitals reported better treatment effectiveness than those in secondary and primary hospitals. These self-reported findings should be interpreted cautiously. In REALISE Asia, about half of adults with asthma had uncontrolled asthma as defined by the Global Initiative for Asthma (GINA), despite available treatment.53 The apparent difference from our findings may be related to disease type, outcome definition, respondent source, follow-up management, and hospital resources. Because our study relied on physician self-assessment rather than patient follow-up data, actual clinical effectiveness may have been overestimated. In addition, the main reasons for treatment failure reported by physicians included inadequate duration of inhaled medication use, irregular inhalation technique, and repeated allergen exposure, which further suggests that improving CVA treatment efficacy requires both optimizing medication selection and strengthening patient education and follow-up management.4,54,55 Although CVA symptoms often improve after treatment, 30%–40% of adults with CVA who do not receive ICS may progress to classic asthma over several years.19 ICS treatment may reduce this risk.20,21 Lower ICS/LABA selection in primary hospitals may therefore affect CVA management in primary care settings, including short-term cough control and long-term prevention of disease progression. Meanwhile, individualized treatment plans are essential for CVA comorbidities, such as allergic rhinitis and gastroesophageal reflux disease, to improve prognosis.56–58 This study found that physicians in tertiary hospitals had a significantly higher awareness of CVA comorbidities and prognostic knowledge than those in primary and secondary hospitals. Such insufficient cognition may also lead to suboptimal treatment outcomes and increased risk of symptom recurrence in CVA patients. This study is the first online survey on CVA diagnosis, treatment, and pulmonary function resource allocation across hospital tiers in China, offering insights for health policy. However, it still had several limitations. First, the self-reported questionnaire data were susceptible to recall and social desirability biases, and were less robust than electronic medical records or objective follow-up data. Pulmonary function test availability and equipment configuration were also based on physician reports and were not independently verified. Second, the WeChat-based online survey may have introduced selection bias. Physicians who were less responsive to online questionnaires may have been underrepresented. Third, because the questionnaire was anonymous and did not collect hospital identifiers, we could not determine whether multiple physicians came from the same hospital or adjust for hospital-level clustering. Finally, this study was conducted only in Zhejiang Province. Therefore, its findings may not represent the situation across the entire country. Future studies should expand to a nationwide scale and include a broader and more representative physician population, objective clinical data, and verified institutional resource data. Prospective or interventional studies are also needed to evaluate whether improved access to bronchial provocation testing can promote standardized CVA diagnosis in lower-tier hospitals, as well as its cost-effectiveness and the value of patient education programs in CVA care. Similar survey frameworks may also be applied to other airway inflammatory diseases, such as allergic rhinitis and chronic rhinosinusitis,59 where diagnostic and treatment resources may vary across care settings. Such evidence may help guide resource allocation and referral strategies. Conclusion This survey reveals differences in CVA management and pulmonary function testing resources across hospital tiers in Zhejiang, China. Compared with tertiary hospitals, primary and secondary hospitals showed lower adherence to guideline-based diagnostic criteria and lower selection of ICS/LABA as first-line treatment. They also had more limited access to pulmonary function tests, especially bronchial provocation testing. The gap between intended and performed bronchial provocation testing suggests that limited test availability may be a potential barrier to standardized CVA diagnosis. These findings may provide a basis for future efforts to strengthen physician training, improve bronchial provocation testing capacity, and optimize referral pathways within the tiered healthcare system. Acknowledgments The authors extend gratitude to Professor Xuefen Wang, Chair of the Zhejiang Provincial Cough Alliance, all members of the Zhejiang Provincial Cough Alliance, and all participating physicians who completed the questionnaire. Funding Statement This study was supported by the Clinical Collaboration Project of Integrated Chinese and Western Medicine for Major Difficult and Complicated Diseases (project code: 2025-ZYYCC-2). Abbreviations CVA, cough variant asthma; ICS, inhaled corticosteroids; LABA, long-acting beta2 agonist; ACT, Asthma Control Test; PFT, pulmonary function test; PCCM, pulmonary and critical care medicine; OR, odds ratio; CI, confidence interval; UACS, upper airway cough syndrome; FeNO, fractional exhaled nitric oxide; PEF, peak expiratory flow; LAMA, long-acting muscarinic antagonist; FEV1, forced expiratory volume in 1 second; FVC, forced vital capacity; GINA, Global Initiative for Asthma; SABA, short-acting beta2 agonist; VAS, visual analog scale. Data Sharing Statement All data generated or analyzed during this study are included in the main text and Supplementary Materials. Individual-level data can be obtained from the corresponding author upon reasonable request. Ethics Approval and Informed Consent This study received ethical approval from the Institutional Review Board for Human Studies of the Second Affiliated Hospital of Zhejiang University School of Medicine (No. I20251625). All participants provided written informed consent, in accordance with the Declaration of Helsinki. Author Contributions All authors made a significant contribution to the work reported, whether that is in the conception, study design, execution, acquisition of data, analysis and interpretation, or in all these areas; took part in drafting, revising or critically reviewing the article; gave final approval of the version to be published; have agreed on the journal to which the article has been submitted; and agree to be accountable for all aspects of the work. Disclosure The authors declare no conflicts of interest in this work. References Associated Data Data Availability Statement All data generated or analyzed during this study are included in the main text and Supplementary Materials. Individual-level data can be obtained from the corresponding author upon reasonable request.