A nationwide, multistage random sampling approach was used across 7 geographic regions in China. In each region, 3 provinces were randomly selected, and 2 municipal-level cities were randomly chosen per province (Table S1 in Multimedia Appendix 1). Following the Hospital Classification Management Measures [35], 4 hospitals were sampled per city: 1 maternal and child health (MCH) hospital, 1 traditional Chinese medicine (TCM) hospital, and 2 general hospitals (including 1 tertiary and 1 secondary).

Articles published between January 1, 2020, and November 30, 2024, were collected via automated web scraping with manual verification. Chinese keyword searches identified BBD-related content (Table S2 in Multimedia Appendix 1). Articles explicitly addressing BBD were included, while those addressing malignancy only, administrative or promotional content, or unrelated topics were excluded (Table S3 in Multimedia Appendix 1).

Data were extracted at both institutional and article levels. Briefly, institutional data—including hospital characteristics, geographic location, bed capacity, and digital engagement—were sourced from official websites and the National Health Commission database. Article-level data—including publication date, view count, and article content—were obtained directly from the sampled WeChat (Tencent Holdings) public accounts. All article view counts were captured on March 16, 2025; views per day were calculated as cumulative views divided by the number of calendar days since publication.

Following established criteria for evaluating online health information [36] and the International Patient Decision Aid Standards (IPDAS) [37], we developed a structured framework encompassing 2 major domains: accessibility and readiness (Figure 1).

Descriptive statistics summarized article and hospital characteristics. Categorical variables were presented as frequencies and percentages. Continuous variables were reported as means (SDs) for normally distributed data or medians (IQR) for skewed data. Chi-square tests were used for categorical variables. Skewed continuous variables were compared using the Kruskal-Wallis test and Bonferroni-corrected Dunn post hoc tests. Interrater reliability for all continuous quality scores (QUEST, Brief DISCERN, and PEMAT-P) was assessed using 2-way random-effects intraclass correlation coefficients for single-rating, absolute agreement (ICC_2,1). For the categorical variables (5C Image Checklist), Cohen kappa was calculated for each domain. For decision relevance content analysis, percentage agreement was used (with an ≥85% threshold).

Linear regression was used to assess temporal trends in publication volume, with year as the independent variable. Given the nested structure of the data (articles clustered within hospitals), linear mixed-effects models with hospital as a random intercept were fitted using the lme4 package in R (version 4.5.0; R Foundation for Statistical Computing). P values were obtained by likelihood ratio tests. A 2-sided P value <.05 was considered statistically significant. NVivo (version 20; QSR International) was used for qualitative data analysis. Datawrapper (Datawrapper GmbH) was used to visualize geographic distribution. Origin 2025 (OriginLab) was used for data visualization.

This study was approved by the ethics committee of Chongqing University Three Gorges Hospital (KLS2025122). As all WPA articles assessed were publicly available and contained no identifying information, the ethics committee granted a waiver of informed consent. Adult patients participated in article evaluation with oral consent. All data were anonymized, and no compensation was provided.

Multistage random sampling yielded 40 city-level units after excluding 2 units: Turpan (Xinjiang) had no independent TCM hospital, and Haibei Prefecture (Qinghai) had no tertiary general hospital at the time of sampling. Of the 160 hospitals expected under the sampling protocol, 158 were included (Figure 2). From these hospitals, 265 WPAs were identified (Figure 2), from which 276,617 articles published between January 2020 and November 2024 were retrieved. After keyword screening, 5526 potentially relevant articles were identified, and 464 BBD-related articles met the inclusion criteria.

The sample comprised 39 tertiary general hospitals, 40 secondary general hospitals, 39 TCM hospitals, and 40 MCH hospitals (Table 1). Overall, 94 (59.5%) hospitals were tertiary-level, and 64 (40.5%) were secondary-level.

In terms of digital engagement, 155 (98.1%) hospitals operated WPAs, achieving full coverage among tertiary general and TCM hospitals. In contrast, only 65.2% (103/158) maintained official websites, ranging from 89.7% (35/39) of tertiary hospitals to 27.5% (11/40) of secondary general hospitals. Certification for internet hospitals was limited (40/158, 25.3%), with the highest proportion among tertiary general hospitals (22/39, 56.4%).

Over the past 5 years, publication volumes on WPAs showed no significant regional differences across China (P>.05; Figure S1 in Multimedia Appendix 1). By hospital type, the median 5-year publication volume ranged from 271 (IQR 83-759) in secondary general hospitals to 743 (IQR 188-1250) in MCH hospitals; secondary general hospitals had significantly lower publication volumes than the other 3 hospital types (Table 1).

Using a nationwide random sampling approach, we evaluated the accessibility and SDM readiness of hospital WPAs in the context of BBD. This study, the first to explore this issue in China, revealed that WPA articles were generally accessible and had the potential to extend SDM beyond clinical encounters. However, the risk-benefit balance was often insufficient, and decision relevance was limited.

SDM in China is still in its early stage, and successful implementation requires culturally appropriate pathways that are customized to fit the local clinical and resource context [47]. Telemedicine can facilitate SDM, and the rapid development of internet hospitals in China provides a new avenue by allowing patients to access information before consultations [48]. In our study, nearly all sampled hospitals operated WPAs, and approximately one-fourth were certified for internet health care, highlighting their commitment to digital service delivery. As part of the expanding digital health ecosystem, public hospitals are increasingly using online platforms such as WPAs to disseminate medical information and engage with patients [49].

Although decision aids are implemented across diverse areas of health care, multiple studies have consistently highlighted significant barriers to their routine use in clinical practice, including time burden and lack of incentives [50]. Effective decision support should convey the right information to the right person in the right format through the right channel and at the appropriate point in the workflow [51]. Incorporating the patient portal into SDM delivery is a promising strategy, as it does not rely on clinicians' initiation [52-54]. This asynchronous SDM also extends decision-making through written or pictorial exchanges across space and time [55,56]. Developing WeChat-based integration tools could therefore represent a pragmatic strategy for embedding SDM into clinical practice.

Despite digital health care platforms providing significant opportunities for SDM, information accessibility remains a critical factor in their effective use. Overall, WPA articles demonstrated acceptable accessibility. Most materials were written at a junior high school level, which was appropriate for the general population, although patients with low health literacy may still face challenges [57]. Visual aids contribute to enhancing patients’ comprehension of health information [58-60]. In our study, visual aids were generally suitable, and optimizing captions and explanatory labels would further strengthen their educational value. From a user-friendliness perspective, materials were generally understandable, but actionability was limited. Although patients could follow the presented information, there was little explicit guidance on what steps to take or how to engage in the decision-making process [40].

These findings align with evidence that hospital WPAs function as hybrid platforms integrating health service delivery with health information dissemination; however, their role has yet to fully extend from providing information to actively supporting patient decision-making [61]. In addition, incorporating multimedia formats—such as images, videos, and interactive tools—can improve accessibility, clarify patient preferences, and foster more effective participation in decision-making [62-64]. Emerging digital technologies, including AI-enabled recommendation systems, may further support personalized decision-making and adherence [65,66].

Effective SDM requires not only the availability of health information but also high-quality and SDM-supportive content [67]. In surgical decision-making for benign tumors, it is crucial for physicians and patients to engage in SDM, with patients receiving balanced information to help them make fully informed choices regarding treatment options [13].

However, balance in treatment-related articles on BBD emerged as a major concern, with only 15.7% (34/216) of articles meeting the criteria for high-quality content. Most articles lacked balanced discussions of potential risks or quality-of-life impacts, both of which are essential for informed decision-making [34]. This aligns with findings from US hospital websites, showing that such imbalanced information could mislead patients and undermine the SDM process [68]. Ensuring that the content patients encounter before consultations is balanced and decision-relevant could make WPAs a practical vehicle for embedding SDM into routine telemedicine.

Patients often rely on clinicians for information, yet inconsistencies arising from institutional practices or individual physician experience may lead to patient confusion [13,69]. Our content analysis revealed wide variation in treatment recommendations, which was primarily reflected in 2 aspects: treatment approaches and surgical decision-making. Most articles presented surgical indications as a combination of tumor characteristics (size and growth rate) and patient-specific factors (symptoms, age, family history, and anxiety), generally supported by literature [4,70-72]. However, the specific thresholds for tumor size that necessitate surgical excision and the definition of “rapidly growing” varied considerably. For instance, the recommended size threshold for fibroadenoma ranged from 2 cm to 3 cm across different regional and international guidelines [70,73,74], resulting in inconsistent information in the WPA articles.

Considerations related to pregnancy were also presented inconsistently. Due to the visible changes in the mammary glands in response to hormonal stimuli, breast masses detected during pregnancy require timely evaluation and individualized management [75,76]. Some of the included WPA articles suggested that surgical management of BBD in the context of pregnancy tends to be proactive in China, which may be partly related to the accessibility of specialized medical services and the anxiety of these patients [77-79]. Women with breast issues often fear the possibility of breast cancer, resulting in unnecessary treatments, patient anxiety, and expenditures [3,80].

Treatment approaches discussed in WeChat articles exhibit a hybrid model that integrates Western medicine and TCM in China [81]. Although VAE remains the dominant surgical approach, emerging treatments such as focused ultrasound also show promising potential [82]. Although TCM has demonstrated benefits in symptom relief and postoperative recovery, its clinical efficacy in treating BBD requires further validation through high-quality studies [83,84]. In the context of BBD, where management options range from observation to intervention, it is important for patients to receive balanced information regarding the risks and benefits of these approaches. Delivering patient-centered care can enhance patient satisfaction and adherence and reduce overtreatment and its adverse impact on quality of life [64,85,86].

Hospital WPAs have the potential to support SDM for BBD, owing to the availability and accessibility of medical information. However, their current SDM readiness is limited, underscoring the need for standardization and quality improvement initiatives to optimize patient-centered decision-making.