This cross-sectional study was conducted from Jan 2024 to April 2024 at the Affiliated Hospital of Inner Mongolia Medical University, focusing on patients with asthma. The study was approved by the Ethics Committee of the author’s Hospital, and informed consent was obtained from all participants.

Inclusion criteria: Patients clinically diagnosed with asthma at the Affiliated Hospital of Inner Mongolia Medical University, aged 18 years or older.

Exclusion criteria: Patients suffering from other systemic diseases such as autoimmune disorders, malignant tumors, or conditions not related to asthma were excluded from the study.

To collect data, convenience sampling was employed to distribute questionnaires to both outpatients and inpatients. Participants were invited to complete the questionnaire by scanning a QR code with their mobile phones on-site. For discharged patients, the QR code was distributed through WeChat groups to facilitate remote participation. All data collection was conducted by trained research personnel at the outpatient and inpatient departments of the Department of Respiratory and Critical Care Medicine. During the survey, some older participants or those with low literacy had difficulty operating smartphones or understanding the questions. To address this issue, our trained staff provided one-on-one assistance, including reading the questions aloud and helping operate the questionnaire system based on the participant’s verbal responses. This ensured accurate and complete data recording.

The required sample size was estimated using the formula for cross-sectional studies (10):

n = Z² × p(1 − p)/d²,

where n is the sample size, Z is the Z-score corresponding to a 95% confidence level (1.96), p is the expected proportion (set conservatively at 0.5 to maximize sample size), and d is the margin of error (set at 0.05). Based on this, the minimum required sample size was n = (1.96² × 0.5 × 0.5)/0.05² = 384.16, rounded up to 385.

A total of 547 valid questionnaires were collected, which exceeds the minimum requirement and ensures adequate statistical power for the study analyses.

The questionnaire design underwent modification after consultation with three experts from the fields of respiratory medicine and otorhinolaryngology. Following their feedback, a pilot study was executed involving 50 randomly selected asthma patients, which yielded a high reliability coefficient of 0.9653. Any issues identified during the pilot phase were quickly resolved to enhance the survey’s effectiveness.

The refined questionnaire encompassed four dimensions, presented in Chinese. These included demographic information with 12 questions; a knowledge dimension consisting of 13 questions, with answers scored as 2 points for a correct response, 1 point for recognition, and 0 points for uncertainty, including some multiple-choice questions that allowed a scoring range from 2 to 18 points; an attitude dimension with 11 questions rated on a 5-point Likert scale, where the intensity of the attitude corresponded to scores ranging from 1 to 5, yielding a total possible score between 11 and 55 points; and a practice dimension comprising 9 questions, also evaluated on a 5-point Likert scale, scoring actions from 1 to 5, with total scores ranging from 9 to 45 points. Scores exceeding 70% of the maximum possible in each section were indicative of adequate knowledge, a positive attitude, and proactive practice (11).

During the formal survey phase, the questionnaire was integrated into the Questionnaire Star platform and disseminated via QR codes. Survey administrators were responsible for promptly verifying the completeness and accuracy of the responses upon submission. To minimize data entry errors and ensure data integrity, all responses were double-entered and each questionnaire was assigned a unique identifier. The research team meticulously reviewed the completeness, internal consistency, and logical coherence of each questionnaire, ensuring the reliability of the collected data.

Data analysis was conducted using R 4.3.2. Continuous data are presented as means and standard deviations (SD) for normally distributed variables, and medians along with the 25th and 75th percentiles for non-normally distributed variables. Categorical data are expressed as counts and percentages [N(%)], describing responses across different demographic characteristics. For continuous variables, a normality test was applied. Comparisons between two groups for normally distributed variables were made using the independent samples t-test, while the Wilcoxon Mann–Whitney test was utilized for non-normally distributed data. For comparisons among three or more groups, ANOVA was conducted for variables with normal distribution and uniform variance, and the Kruskal–Wallis test was employed for non-normally distributed data. Correlation analysis was performed using the Pearson correlation coefficient for normally distributed data and the Spearman correlation coefficient for non-normally distributed data. Regression analyses, both univariate and multivariate, were carried out with dimension scores as dependent variables to investigate the relationships between demographic data and scores. Variables that achieved a p-value of less than 0.1 in univariate analyses were included in the multivariate regression. Classification of mean scores was used for normally distributed data, whereas median scores were used for non-normally distributed data. p-values were reported to three decimal places, considering p < 0.05 as statistically significant. Following the theoretical framework of Knowledge, Attitudes, and Practices (KAP), structural equation modeling (SEM) was used to examine whether attitudes act as mediators between knowledge and practices. Direct and indirect effects were computed and compared. Where model fit thresholds were not met, path analysis was employed for mediation testing. Statistical analyses involving SEM were performed using Stata 18.0.

The study included 547 participants with an average age of 42.66 ± 12.35 years. Of these, 310 (56.67%) were female, 195 (35.65%) had attained a college or bachelor’s degree, 354 (64.72%) resided in urban areas, 299 (54.66%) frequently experienced coughing, 251 (45.89%) had been diagnosed with asthma for no more than 3 years, and 388 (70.93%) had a history of allergic rhinitis. Additionally, 337 (61.61%) of the participants or their primary cohabitants had a history of smoking. Among the 547 participants, the knowledge scores ranged from 2 to 18, with a median of 17, and 63.6% scored above the 70% adequacy threshold. The attitude scores ranged from 11 to 55, with a median of 43, and 72.3% of participants scored above 38.5. The practice scores ranged from 9 to 45, with a median of 33, and 58.4% exceeded the 70% cutoff of 31.5. These results suggest that while general awareness and attitudes are relatively strong, there is more variability in practice behaviors. Analysis of demographic characteristics revealed that knowledge scores varied significantly based on gender (p = 0.003), education level (p < 0.001), occupation (p < 0.001), place of residence (p = 0.014), the presence of immediate family members with asthma (p = 0.030), a history of allergic rhinitis (p = 0.008), social medical insurance status (p < 0.001), and smoking history (p < 0.001). Having an immediate family member with asthma also significantly influenced attitudes (p = 0.002), and educational level, allergic rhinitis history, and smoking history were predictors of varying practice scores (p = 0.040, p = 0.008, and p = 0.044, respectively) (Table 1).

The distribution of knowledge dimensions shown that the three questions with the highest number of participants choosing the “Not clear” option were “The incidence of asthma in allergic rhinitis patients is 4–20 times higher than that in normal individuals. The incidence of asthma in the general population is about 2–5%, while in allergic rhinitis patients, it can be as high as 20–40%.” (K3) with 45.52%, “Allergic rhinitis accompanied by asthma is part of the body’s immune response to recognized foreign substances.” (K7) with 44.%, and “A family history of allergies is a major risk factor for allergic asthma.” (K8) with 24.31% (Table 2).

When it comes to related attitudes, over 64% of participants (n = 351) disagreed or strongly disagreed with the statement “I have enough knowledge to differentiate between allergic rhinitis accompanied by asthma and normal cough” (A9), indicating a significant lack of confidence in symptom distinction. In contrast, 79.2% (n = 434) either agreed or strongly agreed that it is important to understand the difference between asthma and allergic rhinitis for disease management (A2). Furthermore, 92.3% (n = 506) expressed willingness to actively learn how to differentiate between the conditions (A10), suggesting a high potential for educational interventions despite current knowledge gaps (Table 3).

In terms of practice behaviors, 27.61% (n = 151) rarely used different medications to treat allergic rhinitis accompanied by asthma and cough separately (P4), while 38.21% (n = 209) only “sometimes” did so. Similarly, only 5.3% (n = 29) reported always educating their family members and friends about asthma and cough (P9), while 22.12% (n = 121) reported rarely doing so. A total of 18.46% (n = 101) reported “rarely” or “never” recording asthma or cough symptoms for better disease management (P5), suggesting a need to reinforce symptom tracking and patient engagement strategies (Table 4).

Correlation analysis demonstrated that both knowledge and attitude scores were significantly associated with practice. Specifically, knowledge and practice showed a moderate positive correlation (r = 0.2095, p < 0.001), while attitude and practice had a weaker but still significant correlation (r = 0.1420, p < 0.001). No statistically significant correlation was observed between knowledge and attitude (r = 0.0804, p = 0.0601), indicating that knowledge alone may not directly influence attitudes without other mediating factors (Table 5).

Multivariate logistic regression showed that being farmer (OR = 0.09, 95% CI: [0.01,0.87], p = 0.038), being freelancer (OR = 0.06, 95% CI: [0.00,0.59], p = 0.015), being self-employed / businessperson (OR = 0.10, 95% CI: [0.01,0.95], p = 0.045), other occupation (OR = 0.10, 95% CI: [0.01,0.98], p = 0.049), without a family history of asthma (OR = 0.65, 95% CI: [0.44,0.96], p = 0.032), and without history of smoking (OR = 1.67, 95% CI: [1.15,2.43], p = 0.007) were independently associated with poor knowledge. Meanwhile, diagnosed with asthma for 5–10 years (OR = 0.59, 95% CI: [0.37,0.93], p = 0.024) and without a family history of asthma (OR = 0.62, 95% CI: [0.43,0.89], p = 0.011) were independently associated with negative attitude. Furthermore, knowledge score (OR = 1.12, 95% CI: [1.05,1.19], p < 0.001), attitude score (OR = 1.07, 95% CI: [1.00,1.14], p = 0.027), being male (OR = 0.357, 95% CI: [0.136–0.943], p = 0.038) were independently associated with positive practice (Table 6).

The fit indices of the SEM model reached the desired range, indicating good model fit results (RMSEA value: 0.000, SRMR value: 0.015, TLI value: 1.027, and CFI value: 1.000) (Supplementary Table S1), the pathway results showed that smoking history (β = 1.29, p < 0.001) directly affected knowledge. Family history (β = −0.90, p < 0.001) and knowledge (β = 0.15, p < 0.001) directly affected attitude. Knowledge (β = 0.43, p < 0.001) and attitude (β = 0.30, p < 0.001) directly affected practice (Supplementary Table S2 and Figure 1). Indirect effects analyses also showed that smoking history (β = 0.19, p = 0.002) had an indirect effect on attitude. Smoking history (β = 0.47, p < 0.001), family history (β = −0.57, p < 0.001), and knowledge (β = 0.06, p < 0.001) had indirect effects on practice (Supplementary Table S3).