Atopic diseases, also referred to as the atopic march, describe the progression of allergic conditions during infancy and childhood [1]. This progression typically starts with atopic dermatitis or eczema and subsequently progresses to immunoglobulin E–mediated food allergies and allergic rhinitis, eventually leading to asthma if not effectively managed [2,3]. Once initiated, it becomes challenging to halt its course. Epidemiological studies indicate that approximately 75% of children with atopic dermatitis will develop allergic rhinitis, and more than 50% will progress to asthma [1]. In China, the incidence and prevalence of these diseases are showing a rapid upward trend [4]. The prevalence of atopic dermatitis among children aged 1 to 7 years in China was 13% [5]. The overall prevalence of self-reported food allergies among Chinese children aged 0 to 5 years was 4.81% [6]. A meta-analysis estimated that the overall prevalence of allergic rhinitis in Chinese children was 22% [7]. The prevalence of “ever asthma” and physician-diagnosed asthma (16% and 5.3%, respectively) in children aged 3 to 6 years in Shanghai has increased 2-fold over the past 3 decades [8]. The onset and exacerbation of atopic diseases and allergic conditions in children can adversely impact their physical development [9,10], mental health [11], cognitive development [12], and quality of life [13].

Beyond clinical visits and medical treatment, comprehensive caregiver and child education coupled with behavior modification strategies are pivotal for effective atopic disease management. Multidisciplinary interventions have been developed for individual atopic conditions. Among them, some cognitive behavioral therapy approaches targeted at caregiver psychological distress (eg, anxiety, depression, and risk perceptions in food allergy [14-16], asthma [17], and allergic rhinitis [18]), while other educational programs enhanced children’s quality of life and clinical outcomes (eg, reduced disease severity and emergency visits) through caregiver self-efficacy improvement [17,19-21]. However, most existing interventions remain siloed to one specific allergic condition, with limited integration of the atopic march framework despite epidemiological evidence highlighting the high prevalence of comorbid allergic conditions in pediatric populations [22-24]. Given shared pathophysiological mechanisms and management strategies across atopic diseases—including allergen avoidance (eg, dust mite–proof bedding); pharmacotherapy (corticosteroids and antihistamines); barrier protection (nasal saline irrigation and skin moisturizers); immunotherapy; cross-reactivity education; and emergency preparedness (epinephrine autoinjectors) [25,26]—our study posits that unifying these approaches through theory-driven behavioral interventions may synergistically address overlapping pathways. By leveraging the behavior change theory, such integrated strategies could optimize both caregiver and child outcomes while streamlining clinical care for multifactorial allergic conditions.

The protection motivation theory (PMT) was proposed by Rogers in 1975. Since then, the model has been adopted widely to understand, predict, and modify protective behaviors, particularly health-related behaviors [27]. Figure 1 illustrates the structure of PMT. In this model, information from environmental and intrapersonal sources serves as input. The PMT model itself consists of 2 main elements: cognitive mediators and coping models. The cognitive mediators include 6 dimensions and are further divided into 2 categories: threat appraisal and coping appraisal [28]. Threat appraisal involves evaluating unhealthy behaviors or diseases and includes 3 dimensions: susceptibility (how susceptible an individual feels to the threat), severity (how serious the threat is perceived to be), and internal or external rewards (the benefits of not taking protective action) [27,29]. Coping appraisal involves assessing health behaviors and includes 3 dimensions: self-efficacy (confidence in one’s ability to carry out the adaptive response); response efficacy (belief in the effectiveness of the adaptive response); and response cost (the perceived costs associated with taking the adaptive response, such as time, effort, or money) [30]. Coping modes, on the other hand, represent the behavioral outcomes and can be either adaptive responses (intentions to adopt healthy behaviors) or maladaptive responses (intentions to adopt unhealthy behaviors) [29]. These 6 dimensions within the 2 cognitive mediators form the core constructs of PMT.

Over the past decades, the PMT has been widely applied across diverse health domains with notable efficacy. Its applications extend beyond adult health behavior modification, such as skin cancer prevention [31], prenatal environmental risk education [32], and lifestyle management in military personnel with type 2 diabetes [33]. The PMT has also been effectively used in caregiver-mediated interventions for pediatric populations. These include improving children’s asthma symptoms [34]; oral health [35]; enhancing treatment adherence in muscular dystrophy [36]; promoting COVID-19 vaccination uptake [37]; and mitigating risks related to sexual abuse [38], pornography exposure [39], online hazards [40], and secondhand smoke [41]. Given the PMT’s validated effectiveness in child health management across these applications, we propose it as a theoretically grounded framework for guiding our online cognitive behavioral intervention for pediatric atopic diseases.

Moreover, online health interventions demonstrate advantages over traditional in-person approaches, including enhanced convenience, accessibility, and cost-effectiveness and reduced social stigma through anonymity [42-45]. These online programs have been successfully applied in caregiver-mediated pediatric interventions, spanning unintentional injury prevention [46], breastfeeding support for preterm infants [47], and cancer symptom management [48] as well as applications targeting pediatric atopic diseases (eg, reducing disease severity [19], improving treatment adherence [49,50], and alleviating caregiver psychological distress [18]). These findings collectively highlight the potential of digital behavioral interventions to address multidimensional challenges in atopic disease management.

The primary aim of this study was to conduct a quasi-experimental study to evaluate the effects of a 4-week PMT-based cognitive behavioral online (PMT-CBO intervention) in improving caregivers’ engagement in managing their children’s atopic diseases. Specifically, this study sought to determine the extent to which the intervention improves caregivers’ protection motivation, intentions, and behaviors. The secondary aim of this study was to further investigate the impacts of the PMT-CBO intervention on children’s atopic diseases outcomes, including children’s medication adherence and subjective disease symptoms. We hypothesized that participants assigned to the PMT-CBO group would have significantly greater improvements in both the primary and secondary outcomes compared to those assigned to the control group, who only received routine verbal education during pediatric outpatient visits.

This study, using a quasi-experimental design, was conducted in Hangzhou, China. Three health care institutions with high outpatient volumes were purposively selected from 3 districts of Hangzhou: a general hospital in Yuhang District (location Y), an integrated hospital in Xihu District (location X), and a general hospital in Gongshu District (location G). Locations Y and X were assigned to the PMT-CBO group, while location G served as the control group. This design intentionally substituted individual randomization with institutional-level assignment to reflect real-world implementation constraints and to control contamination risks through spatial segregation. Baseline equivalence testing confirmed comparable socioeconomic profiles across the groups.

The sample size was calculated based on the primary outcome indicator, the protective motivation score of caregivers, using the formula for comparing the means of 2 independent groups:

Where σ is the estimated population SD, δ represents the mean difference between groups, Z_α/2 is 1.96 for the significance level α of .05, and Z_β is 1.28 for a test power of 80%. On the basis of prior cross-sectional survey data from the research group, the estimated overall protection motivation score was 3.50 (SD 0.47), yielding σ=0.47. Considering an anticipated increase of at least 0.30 points in the score (δ=0.30) after the PMT-CBO intervention, the estimated sample size was 52 for each group. Accounting for an estimated loss rate of 20%, the sample size was adjusted to a total of 126 individuals, with at least 63 in each group.

Figure 2 illustrates the full study flow diagram. Recruitment was conducted between May and December 2023 during pediatric outpatient clinic visits for children with physician-diagnosed allergic diseases. Eligible families were approached by attending pediatricians to assess parental willingness to participate. Enrollment was voluntary, with caregivers providing written informed consent after receiving detailed study information. Caregivers were eligible if they (1) had at least one child diagnosed with atopic diseases, aged 0 to 14 years; (2) were the primary caregiver living with the child; (3) had functional literacy and smartphone proficiency; (4) and voluntarily participated and signed the consent form. Exclusion criteria included (1) participation in other cognitive behavioral intervention trials and (2) the child having severe physical or mental illnesses or complications requiring surgical treatment.

A total of 243 participants were enrolled and completed the baseline assessment. Of these, 127 (52.3%) from locations Y and X were allocated to the PMT-CBO group and 116 (47.7%) from location G were assigned to the control group.

The research tools, procedures, and consent forms were approved by the Ethics Review Committee of the School of Public Health, Hangzhou Normal University (approval no. 20230002). All participants signed the consent form and received a US $1.50 appreciation gift upon enrollment. All personal data in the study have been anonymized during the data organization and statistical analysis stages to ensure participant privacy.

The normality tests were conducted on continuous variables using the Shapiro-Wilk test for samples with n≤50 or the Kolmogorov-Smirnov test for samples with n>50. The normally distributed continuous variables were presented as mean (SD), nonnormally distributed continuous variables were expressed as median (IQR), while categorical variables and ordinal variables were reported as counts (percentages). Intergroup differences were evaluated using the independent t tests for normally distributed continuous variables, Mann-Whitney U tests for nonnormally distributed continuous or ordinal variables, and chi-square or Fisher exact test for categorical variables. Intragroup differences between pre- and postintervention measures were examined with paired 2-tailed t tests for normally distributed variables and Wilcoxon signed-rank tests for nonnormally distributed continuous variables. The statistical significance level for all tests was set at P<.05 (2-tailed).

Given that the evaluation metrics of protection motivation were nonnormally distributed, and to control for potential confounders that might influence the evaluation of the intervention’s effect, the study used a multivariable generalized estimating equation (GEE) model with an exchangeable correlation structure adjusting for impactful demographic characteristics. In this model, the various dimensions of protective motivation and the overall PMT score were treated as individual outcome variables. The group assignment (experimental vs control) and time point (after the intervention vs before the intervention) served as the exposure variables. To capture how the potential intervention effect might change with time, an interaction term between time and group was also incorporated. Furthermore, demographic and disease-related factors were included as potential confounders, including the child’s age, sex, caregiver identity, household registration status, parental history of atopic diseases, and the child’s birth order within the family. These confounders (eg, child’s age, sex, parental history of atopic diseases, child’s birth order within the family [61,62], and the caregiver identity [6,63]) were selected based on prior literature documenting risk factors for pediatric atopic disease prevention and control. Household registration status was used as a proxy for environmental determinants and health care accessibility [4,64]. On the basis of the data distribution characteristics, the Gamma GEE with a log link function and type III analysis was used to analyze nonnormally distributed continuous variables after the 4-week intervention. Statistical analyses were performed with SPSS (version 27.0; IBM Corp) and R (version 4.3.1; R Foundation for Statistical Computing).

After a 4-week intervention, a total of 116 participants from the experimental group and 111 participants from the control group completed the posttest questionnaire, with a completion rate of 91.3% (116/127) and 95.7% (111/116), respectively (Figure 2). A chi-square test was conducted to investigate whether there was a significant difference between the control group and the experimental group in terms of lost visits during the intervention process (χ²₁=0.1; P=.80). These results showed that the number of participants was still balanced between the 2 groups after the intervention, with no significant difference in their rates of lost visits. All data analysis were performed using intention-to-treat analysis to reflect the true effects of the intervention under the quasi-experimental design and to minimize the influence of selection bias on the study conclusions (Table 1).

The baseline characteristics of the children recruited in the study and their subjective disease symptoms and medication adherence status are summarized in Tables 2 and 3. The median age of the allergic children was 4.00 (IQR 3.00-7.00) years, average height was 109.79 (SD 20.52) cm, and median weight was 18.50 (IQR 15.00-25.00) kg, with 63% (153/243) of them being boys. Among the types of atopic diseases, most children (103/243, 42.4%) were affected by allergic rhinitis, while a substantial proportion (103/243, 42.4%) were also reported to have more than 1 atopic disease. Regarding self-reported disease symptoms and medication adherence status, the median or mean scores for POEM, FAQL-PB, AR-SRS, and C-ACT were 7.00 (IQR 5.00-9.00), 2.00 (IQR 1.88-2.35), 4.00 (IQR 3.00-5.00), and 19.46 (SD 3.48), respectively, and the median score for GMAS was 25.00 (IQR 24.00-26.00), indicating relatively poor symptom control and a low quality of life for caregivers.

The baseline characteristics of the caregivers are presented in Table 4. Among the caregivers, most of the participants were mothers, comprising 76.5% (186/243) of the sample. At baseline, the caregivers’ protective motivation regarding their children’s atopic diseases was assessed using a developed and validated PMT scale. This scale measured 6 separate dimensions of protection motivation, with each dimension scored on a scale ranging from 0 to 100. The median scores for the 6 dimensions were as follows: susceptibility, 83.33 (IQR 75.00-91.67); severity, 75.00 (IQR 75.00-83.33); self-efficacy, 58.33 (IQR 50.00-66.67); response efficacy, 78.57 (IQR 75.00-85.71); response cost, 41.67 (IQR 33.33-50.00); and internal or external rewards, 41.67 (IQR 33.33-50.00). The overall PMT score achieved a median score of 65.91 (IQR 61.36-69.32), far from the maximum score of 100. At baseline, no substantial differences were observed between the intervention and control groups concerning caregivers’ demographics, protective motivation, or children’s demographics, severity and types of atopic diseases, frequency of attacks, subjective symptoms, and medication adherence.

First, both changes of PMT scores within the PMT-CBO group before and after the intervention and the changes within the control group were calculated and are presented in Figure 3 and Multimedia Appendix 3, which are illustrated as intragroup differences. It demonstrated that, following the intervention, the scores across all 6 dimensions of protective motivation, as well as the overall PMT score, significantly increased in the PMT-CBO group compared to the preintervention (P<.001). Among them, the scores for the dimensions of severity (median 8.33, IQR 0-16.67), self-efficacy (median 8.33, IQR 0-16.67), response cost (median 8.33, IQR 0-16.67), and internal or external rewards (median 8.33, IQR 0-16.66) increased substantially (see Multimedia Appendix 3 for details). This finding indicates a substantial enhancement in the protective motivation of caregivers in the PMT-CBO group after the intervention. Moreover, in the control group, postintervention scores were significantly higher than preintervention scores across most PMT dimensions except for the response cost dimension (P<.05; Figure 3 and Multimedia Appendix 3). Second, a comparison analysis of the score difference between the 2 groups before and after the intervention, illustrated as intergroup difference in Figure 3, revealed that the PMT-CBO group exhibited significantly greater improvements than the control group in both the overall PMT score (median 5.68, IQR 3.41-12.50 vs median 2.27, IQR 0-5.68) and most of individual dimensions (P<.001), excluding the dimensions of susceptibility and internal or external rewards.

The results of the multivariable GEE models to assess changes in the dimensions of protective motivation and the overall score from before to after the intervention are presented in Table 5. It indicated that, according to the effect of the group-time interaction term (β=.027; P=.03), the PMT-CBO intervention based on the PMT implemented in the experimental group showed a significant positive impact on the susceptibility dimension compared to the routine intervention received by the control group. Similarly, the experimental group demonstrated significantly greater improvements in the dimensions of severity, self-efficacy, response efficacy, and response cost, as well as in the overall protective motivation score (P<.001), with response cost obtaining the highest effect of group-time interaction (β=.231). These findings suggested that, even after controlling for potential confounding factors, the cognitive behavioral online intervention effectively enhanced caregivers’ knowledge of and proactive engagement in managing pediatric atopic diseases. However, in the specific dimension of internal or external rewards, no significant interaction effect was observed between time and group (β=.005; P=.90), indicating that the improvement in this dimension in the experimental group was not substantially different from that in the control group.

In this study, we also assessed both intra- and intergroup changes in caregivers’ intentions to prevent and control their child’s atopic conditions. As shown in Table 6, after 4 weeks of intervention, caregivers in the PMT-CBO group exhibited significantly higher levels of proactivity and improvement compared to the control group across most key indicators of prevention and control intentions (P<.05). Specifically, significant improvements were observed in the intentions to prevent their child’s allergic diseases in general (Z=–2.665; P=.008); use corticosteroid-based antiallergic medications (Z=–3.442; P<.001); control allergens regularly (Z=–3.025; P=.002); receive health education related to allergic diseases (Z=–2.403; P=.02); seek medical treatment promptly and adhere to prescribed medications (Z=–2.275; P=.02); and attend follow-up appointments regularly (Z=–2.202; P=.03). However, no significant improvement was observed in caregivers’ intentions to monitor their child’s condition regularly when compared to the control group (Z=–1.873; P=.06).

In terms of behavioral outcomes related to the prevention and control of atopic diseases, caregivers in the PMT-CBO group reported lower incidences of their child’s atopic diseases due to allergen exposure compared to those in the control group both before (Z=–2.429; P=.02) and after the 4-week intervention (Z=–2.396; P=.02; Multimedia Appendix 4). In addition, a significantly higher percentage of caregivers in the PMT-CBO group attended health education sessions compared to the control group (Z=–2.149; P=.03), likely as a direct result of the PMT-CBO intervention. Within the PMT-CBO group, significant improvements were observed after the intervention in behaviors such as attending follow-up appointments regularly (Z=–2.371; P=.02) and adhering to prescribed medication regimens (Z=–2.077; P=.04). However, no significant inter- or intragroup differences were observed for other prevention and control behaviors, such as adopting general preventive measures for atopic diseases (Z=–1.016; P=.31) or monitoring the child’s condition regularly (Z=–0.888; P=.37). These findings may be attributed to the relatively short duration of the intervention, which might have been insufficient to elicit substantial behavioral change.

Within the PMT-CBO group, medication adherence as measured by GMAS improved significantly after the intervention compared to the preintervention period (Z=–7.970; P<.001; Figure 4 and Multimedia Appendix 5). The median difference in GMAS scores between before and after the intervention for the PMT-CBO group was 1.00 (IQR 0-2.00). In contrast, no significant improvement in medication adherence was observed in the control group following routine care and verbal education (Z=–1.673; P=.09; see Multimedia Appendix 5 for details). Furthermore, when comparing the 2 groups, the improvement in children’s medication adherence scores in the PMT-CBO group was significantly greater than that in the control group (Z=–4.457; P<.001). These findings highlight the significant effect of the PMT-CBO intervention in improving children’s medication adherence through caregiver engagement.

The results presented in Figure 5 indicate that, compared to the preintervention period, children in the PMT-CBO group showed a significant decline in POEM, FAQL-PB, and AR-SRS scores following the intervention (all P values <.01). For instance, the median difference in POEM scores between before and after the intervention for the PMT-CBO group was 2.00 (IQR 2.00-3.00). These reflected improvements in children’s eczema symptoms, caregiving burden for caregivers of children with food allergies, and children’s allergic rhinitis symptoms. Meanwhile, the C-ACT score significantly increased in the PMT-CBO group (t₄₆=–5.582; P<.001), with a median score difference of 1.00 (IQR 0-2.00), indicating a marked improvement in asthma control in these children. In the control group, significant improvements were observed only in FAQL-PB (Z=–4.307; P<.001) and C-ACT (t₂₀=–3.440; P=.003) scores (see Multimedia Appendix 6 for details).

More importantly, in terms of intergroup differences, the PMT-CBO group exhibited significantly greater improvements in POEM (Z=–3.112; P=.002) and AR-SRS (Z=–3.277; P<.001) scores compared to the control group (see Figure 5 and Multimedia Appendix 6 for details). However, no significant intergroup differences were observed for changes in FAQL-PB (Z=–1.693, P=.09) and C-ACT scores (Z=–.830, P=.41). These findings suggested that the PMT-CBO intervention effectively improved eczema and allergic rhinitis symptoms in children.

This study demonstrated that the 4-week PMT-CBO intervention yielded significant intragroup improvements in caregivers’ overall PMT scores and all PMT dimensions in pre-post comparisons. Intergroup comparisons confirmed the superior efficacy of PMT-CBO group over controls in most PMT dimensions, except for susceptibility and internal or external reward dimensions. Furthermore, caregivers receiving the PMT-CBO intervention displayed better outcomes in key indicators of atopic prevention and control intentions, alongside partial improvements in relevant behavioral measures related to disease management. Regarding child-level outcomes, children in the PMT-CBO group demonstrated improved intergroup medication adherence and clinically meaningful alleviation of eczema and allergic rhinitis symptoms compared to the control group, while no significant differences emerged in asthma control metrics or caregiver life burden associated with food allergy management.

Among 6 PMT dimensions, self-efficacy and response efficacy exhibited the most pronounced intergroup enhancements in our study, which was aligned with previous PMT-based studies of caregivers of children with asthma [34]. The marked elevation in self-efficacy and response efficacy likely stemmed from the intervention’s skill-building components, such as structured guidance on allergen avoidance, medication adherence, and daily care routine and educational content emphasizing the benefits of adherence. This was consistent with findings from the trial by Lima et al [20], where education content contains topics such as the following: “Let’s learn how to reduce asthma triggers”; “Controlled asthma improves health and well-being” in educational booklets significantly enhanced caregivers’ self-efficacy and response efficacy for asthma management. In addition, other studies with similar intervention content significantly improved the self-efficacy and response efficacy of other caregivers of children with eczema and food allergy [65,66]. In contrast, susceptibility perceptions showed no significant intergroup difference, potentially due to high baseline awareness of genetic risks in families with atopic histories [67]. Saturation in preexisting risk cognition may have limited further elevation. Meanwhile, limited intergroup changes were observed in internal or external rewards, possibly because caregivers inherently viewed disease management as an obligatory responsibility rather than an optional behavior meriting external incentives [68]. Future interventions should be tailored to specific characteristics (eg, family history) of the target population, with adjustments in content of each PMT dimension to achieve a more balanced and comprehensive enhancement in parental motivation.

Our PMT-CBO intervention elicited notable improvements in caregivers’ behavioral intentions but limited behavioral changes, suggesting a time lag between cognition and action. The observed discrepancy between enhanced behavioral intentions and only modest behavioral adjustments may indicate that the effects of cognitive improvements require a longer duration to manifest as observable changes in behavior [69]. To address this, it is recommended that future interventions not only continue to focus on enhancing cognitive mediators but also incorporate strategies from the theory of planned behavior (TPB) to reinforce the intention-behavior link [70]. Integrating TPB principles could help caregivers in developing concrete action plans and setting realistic behavioral goals [71,72]. Moreover, external environmental factors appear to play a moderating role in shaping health behaviors. For instance, limited access to health care resources and the burden of economic costs may further impede the adoption of improved disease management behaviors despite increased intentions [73,74]. Such external constraints emphasize the need for multilevel environmental support strategies and ecological frameworks for health promotion.

The multidimensional integrated intervention in this study demonstrated significant synergistic advantages in managing allergic diseases by addressing shared pathophysiological pathways and behavioral mechanisms. Specifically, epithelial barrier dysfunction and Th2-mediated inflammation may be alleviated by our unified interventions, such as instructions of daily moisturizing routines and environmental trigger avoidance, which collectively disrupt the progress of “atopic march.” This aligns with recent evidence showing that early skin barrier enhancement in infants [75-77] or allergen reduction [78-80] can lower IgE sensitization risks and reduce comorbid allergies, supporting the clinical value of holistic approaches over siloed disease management. Furthermore, the cross-disease enhancement of medication adherence may be explained by skill transferability. Caregivers who mastered topical steroid application for eczema could seamlessly adapt these techniques to rhinitis nasal sprays, reinforcing self-efficacy through competency generalization. Positive feedback loops emerged as symptom relief in one condition (eg, reduced eczema pruritus) increased caregivers’ confidence in managing comorbid allergies, creating a “virtuous cycle” of proactive care [81,82]. These findings advocate for standardized clinical protocols integrating barrier maintenance, environmental controls, and skill-based medication training as foundational components of pediatric allergy prevention programs.

Despite these synergies, disease-specific challenges persist, particularly in asthma and food allergy management. The suboptimal asthma control outcomes in this study may reflect the unique temporal and technical demands of this condition—unpredictable exacerbations disrupt caregivers’ threat appraisal consistency, while an improper inhaler technique negates controller medication efficacy. Prior research indicates that 60% to 80% of caregivers commit critical inhaler errors [83,84] and the number decreased significantly after standard training [85,86], underscoring the need for condition-specific reinforcement. Similarly, the unmitigated burden of food allergy management exposes a conflict between uncontrollable social dining risks and hypervigilant parenting behaviors. Caregivers tended to avoid social gatherings (eg, birthday parties) due to allergy anxieties or worries, paradoxically increasing children’s social isolation and loneliness risks [87-89]. To address these disparities, modular intervention add-ons could be developed in the future, such as gamified inhaler trainers with real-time artificial intelligence feedback [90,91], or virtual reality simulations of restaurants and food to desensitize families to controlled allergen exposures [92,93]. Embedding these condition-specific modules within the current core PMT framework would preserve integration benefits while accommodating disease heterogeneity—a critical step toward precision and individualized allergy care.

Our findings revealed the multidimensional complexity of pediatric allergy management. On the basis of this PMT trial, we proposed a stepped-care program combining core PMT modules, characteristic or condition-specific plug-ins, and community support systems to balance scalability with personalization in the future. PMT-based core modules would deliver universal behavioral principles (eg, trigger identification, emergency protocols, and adherence improvement), as validated in our study. Meanwhile, characteristic- or condition-specific plug-ins could tailor content to dynamic needs, such as notifying personalized genetic risk feedback, intensifying asthma action plans during pollen seasons [94], or informing food allergy labeling in shopping and dining scenarios [95]. Moreover, merging the PMT with the TPB or other implementation intention techniques (eg, “if-then” planning for asthma exacerbations) may strengthen the intention-behavior pathway [96,97], as shown in children’s oral health programs [98]. To further enhance PMT’s applicability to atopic management, theoretical integrations should also be considered. While the PMT effectively addresses individual-level cognition, its limited emphasis on social and environmental determinants constrains behavior change in real-world settings. Therefore, incorporating constructs from the socio-ecological model may help address these multilevel barriers, such as health care system support (eg, regular allergy testing access) and community or school resources (eg, school allergy-free policies) [99-101].

Our study explored the application of PMT to develop integrated caregiver-mediated interventions for children with atopic diseases, novelly considering the progression of the atopic march and comorbidity patterns through a cognitive behavioral framework. We developed multidisciplinary intervention resources (eg, trigger identification and emergency preparedness) to address key and universal challenges in atopic disease management, ultimately improving caregivers’ protection motivation, behavioral intentions, and children’s clinical outcomes (eg, eczema severity reduction and medication adherence). In addition, the replicable and scalable online intervention package was embedded within a WeChat mini-program. The WeChat platform is accessible to >1.3 billion users and has no installation barriers in China, ensuring 24-7 availability and continuity of education [102]. The WeChat group chats further facilitated communication between caregivers, the research team, and health care professionals, enhancing intervention effectiveness by providing timely and tailored support.

There are several limitations in this study. First, this study may have limited generalizability as participants were recruited from urban, educated populations in more-developed eastern China [64,103]. Future research should include more diverse samples from rural or less-developed areas for broader representativeness. Second, evaluation of the study relied on online self-reported questionnaires or scales. While these tools demonstrated good reliability and validity, they may be susceptible to subjective bias. Future studies should integrate objective measures and economic indicators to strengthen validity. Third, the 4-week intervention duration was relatively short, necessitating extended follow-up to evaluate its long-term effects on caregivers’ behaviors and children’s health outcomes. Extended follow-ups were needed to evaluate its long-term effects on caregivers’ behaviors and children’s health outcomes. Fourth, the PMT-CBO group showed slightly lower completion rate than the control group (116/127, 91.3% vs 111/116, 95.7%), possibly due to the intervention’s complexity. Future iterative usability testing and optimized incentive strategies may help enhance adherence and completion rate.

This study demonstrates that a 4-week PMT-CBO intervention effectively enhances caregivers’ motivation, intentions, and partial behaviors in managing pediatric atopic diseases, while improving children’s medication adherence and alleviating eczema and allergic rhinitis symptoms. By integrating the intervention package into a scalable WeChat-based platform, the intervention achieved high accessibility and engagement (>90%), offering a replicable model for digital allergy education. However, limited effects on asthma control and food allergy–related caregiver burdens highlight the need for condition-specific adaptations, such as artificial intelligence–enhanced inhaler training or food allergen risk simulations. Future work should prioritize socioeconomic diversity, objective outcome measures, and ecological frameworks to bridge intention-behavior gaps and promote equitable, precision allergy care.