Guided by a phenomenological perspective and using qualitative research methods, this study aims to explore in depth the experiences and feelings of Chinese patients with breast cancer as they search for web-based health information during their disease journey [22]. Data were collected through semistructured face-to-face in-depth interviews. This study drew upon Lazarus and Folkman’s [15] stress-coping theory in designing the interview guide. Following thematic analysis, researchers used this theoretical framework to interpret patients’ cognitive evaluations and coping strategies within their web-based information-seeking behaviors, thereby enhancing the theoretical significance of the findings. Originally proposed in 1984, this classic psychological theory explains how individuals assess and respond to stressful events through a dynamic regulatory process involving person–environment interactions. The study reporting followed the COREQ (Consolidated Criteria for Reporting Qualitative Research Checklist; Checklist 1) [23].

From September 2024 to May 2025, participants were recruited through a combination of convenience and purposive sampling in 3 breast cancer inpatient wards at Hunan Cancer Hospital in China. Hunan Province is a typical province in central China, comprising 14 prefecture-level cities. Its diversity in language and culture, urban-rural structure, health care resources, economic development, and residents’ digital literacy provides a suitable setting for examining the contextual variations in information-seeking behaviors [24].

Two female researchers (LJH and ZD) were Advanced Practice Nurses in breast oncology with training in breast disease management and qualitative interviewing. The researcher has experience practicing in participants’ workplaces and is responsible for participant screening and interview scheduling. First, based on convenience sampling principles, the researcher reviewed daily inpatient records and made preliminary assessments of eligibility based on medical documentation. Subsequently, clinicians and charge nurses assisted in notifying eligible patients. Researchers then conducted face-to-face interactions with patients, explaining the study objectives, procedures, timeline, and privacy safeguards to invite participation. To enrich the data dimensions, the research team purposefully included patients of varying ages, educational levels, disease stages, duration since diagnosis, and urban/rural residence after initial recruitment, continuing until data saturation was achieved.

Participants were female patients diagnosed with breast cancer who had experience in searching for information on the internet. Inclusion criteria included the following: (1) aged 18 years and older; (2) diagnosed with breast cancer; (3) able to express themselves clearly, willing to participate in the study, and sign an informed consent form; and (4) previously actively searched for health information on the internet or mobile platforms during the disease. Exclusion criteria included the following: (1) comorbid serious mental illness affecting the interview communication; (2) those who refused to record or were unable to complete the interview.

Data were collected through semistructured, in-depth interviews. Based on the study aims, a review of relevant literature, and the theoretical framework of stress and coping, the interview guide was developed collaboratively by the research team members (LJH and QCL) with guidance from a qualitative research expert (LXY). To enhance clarity, exploratory questions were incorporated into the primary interview questions to better capture the depth and complexity of participants’ experiences, as shown in Table 1. Two participants who met the inclusion criteria were invited for a pilot interview. Based on the feedback from the pilot participants, the wording and sequence of the interview questions were adjusted, and the exploratory questions were also optimized.

The interview began after obtaining the participants’ verbal and written informed consent. The entire process was audio-recorded after obtaining their consent, and at the same time, the researchers took field notes to capture nonverbal information and key immediate thoughts. During the interview, the researcher flexibly adjusted the order and depth of questions according to the participants’ expressions and responses and encouraged patients to describe their information-seeking experience in-depth and truthfully. The average duration of the interviews is approximately 45 minutes (range 20‐60 min). All interviews were conducted in quiet hospital spaces designated for patient–clinician communication to ensure privacy and comfort. To minimize interviewer bias and ensure the authenticity of participants’ expressions, the researcher adopted a neutral and nonevaluative stance during the interviews, avoided leading or suggestive questions, and respected each participant’s subjective experiences. When no new topics are generated, data collection is halted [25]. When no new significant themes emerge after analyzing 2 consecutive rounds of interviews, data saturation is achieved, and sampling ceases.

The interview recordings were transcribed verbatim into Chinese text by the researcher (LJH) within 24 hours of completion and verified for accuracy by the researcher (ZD). Transcribed texts are managed using Excel, with each record containing participant ID, raw text, open coding, thematic categorization, and researcher notes. Excel files are uniformly stored in a controlled environment and encrypted to ensure participant information security and privacy.

Using thematic analysis as the methodological basis for theme summarization [26]. Two researchers (LJH and ZD) independently conducted the initial coding and then discussed and reached a consensus. In case of disagreement, a third-party expert (LXF) was invited to participate in the discussion to ensure the reliability and authenticity of the analysis. After coding completion, the researcher (LJH) categorized and merged relevant codes to form preliminary themes. These themes underwent discussion and review by all researchers to ensure alignment with the original data. Subsequently, themes were defined and named, with descriptive explanations drafted for each. Finally, themes were linked to corresponding original data citations. The researchers provided the preliminary analysis results to some participants for feedback and solicited their opinions to ensure the authenticity and representativeness of the analysis. To ensure the accuracy of language conversion, all citations and theme descriptions intended for publication were translated from the original Chinese into English by professional bilingual personnel with medical backgrounds. Subsequently, all members of the research team conducted back-translation and concept [27].

Throughout the research process, the team maintained a reflective stance to enhance transparency, credibility, and rigor. Given that the interviewers had professional backgrounds in the breast department, face-to-face semistructured interviews fostered trust, enabling participants to be more willing to share their personal experiences. However, this also risked leading researchers to overinterpret participants’ perspectives during interviews, potentially overlooking subtle variations in their experiences. Therefore, researchers maintained reflective journals before and after each interview, documenting their preconceptions, emotional responses, interview interactions, and any subjective impressions that might influence interpretation. These records were repeatedly reviewed during subsequent analysis. During data analysis, researchers repeatedly returned to original recordings to ensure initial coding stemmed from participants’ authentic expressions. Researchers took particular care to separate clinical experience from the analytical process, avoiding the substitution of professional judgments for meanings presented in participants’ narratives. Throughout theme development, the research team continuously challenged, compared, and refined interpretations through ongoing peer discussions and reflective dialogues. This multiperspective examination of coding consistency and thematic robustness enhanced the credibility and confirmation of research findings. Additionally, the team deliberately revisited earlier hypotheses and analytical pathways during the later stages of analysis [28].

The study was approved by the Hunan Cancer Hospital Medical Ethics Review Committee (number SBQLL-2024‐180). All participants signed written informed consent forms prior to the commencement of face-to-face interviews, ensuring their participation was voluntary, information was anonymized, and data were used solely for this study. Furthermore, no financial compensation was provided to any participants. All raw interview data and participant information underwent anonymization and were stored on an encrypted cloud storage drive to ensure confidentiality.

A total of 18 female patients with breast cancer participated, originating from Wuhan, Hubei Province (n=1), Jiamusi City, Heilongjiang Province (n=1), and 9 prefecture-level cities in Hunan Province (Changsha, Yiyang, Changde, Hengyang, Loudi, Zhuzhou, Shaoyang, Yongzhou, and Yueyang) (n=16). All the participants were of Han ethnicity. All patients with breast cancer are currently undergoing treatment in the hospital. Except for one participant who had bilateral breast cancer, all the other participants had unilateral breast cancer. The time of cancer diagnosis for the participants ranged from 1 month to 20 years. The most commonly used digital search platform by the participants was Baidu. The general information of the participants is shown in Table 2.

This study identified three core themes and 11 subthemes of OHIS in patients with breast cancer through thematic analysis: (1) the driving forces of information search were emotion-driven and problem-oriented; (2) cognitive judgments in the information fog included the interweaving of multiple channels of information, assessment of information authenticity and applicability, cognitive confusion under information conflicts, and the construction of information meaning; (3) adaptive strategies under the pressure of network information encompassed information conversion into action, emotion regulation, support network construction, and expectation acceptance and adjustment. Overall, the OHIS behaviors of patients with breast cancer present a dynamic process driven by motivation, cognitive processing, and adaptive strategies. The overall coding framework, including the hierarchical relationship between themes and subthemes, is shown in Figure 1.

The findings of this study authentically reflect the web-based information-seeking experiences of patients with breast cancer, which can be summarized under 3 core themes: drivers of information seeking, cognitive judgments amid information fog, and adaptive mechanisms under web-based information pressure. These themes are not isolated categories but rather depict a dynamic psychosocial process: how patients navigate a challenging digital information environment from their initial needs to ultimately develop personalized coping strategies.

This study found that information-seeking behavior among patients with breast cancer is driven by both emotional and problem-solving needs. The anxiety and uncertainty brought by a diagnosis drive emotional information-seeking centered on seeking reassurance, while simultaneously prompting problem-solving searches to understand the disease and participate in decision-making. Consistent with the classification of web-based help-seeking behaviors of patients with cancer in previous studies, the study by Feng et al [29] mainly focused on web-based health communities and categorized patients’ help-seeking behaviors into informational and emotional support needs. This study further expands this perspective by extending the scope of analysis from help-seeking behaviors in web-based community interactions to a broader context of OHIS. Notably, these 2 drivers do not switch off and on simply; they often coexist and reinforce each other. This finding resonates with research by Liu et al [30], based on a dual-factor model, which confirmed that pediatric patients’ OHISs are simultaneously influenced by complex factors—both facilitators (such as perceived stress) and inhibitors (such as learned helplessness), revealing the inherent tension within behavioral drivers. In the Chinese context, this driving force is further moderated by the family system. Phenomena observed in this study, such as spousal discouragement and proxy searching by children, align with Feng’s [31] findings that Chinese family communication patterns influence information-seeking behaviors undertaken for family members or oneself. This highlights how, within collectivist cultures, disease management is often viewed as a shared family unit event.

This study reveals patients’ cognitive judgment processes within uncertain information environments. Patients develop strategies such as cross-verifying multiple channels, tracing information sources, and assessing applicability to construct personal cognitive safety nets amid information fog. The core paradox of this process lies in the dual role of information: while it may enhance patients’ sense of control over their illness, excessive web-based information, poor quality content, and conflicting viewpoints can also trigger decision-making difficulties, diminished trust in information, and persistent anxiety [32,33]. This study particularly highlights the critical role of doctor-patient communication within the information fog. When patients conceal their web-based doubts for fear of challenging medical authority, conflicting information becomes internalized as a psychological burden. This finding strongly corroborates the study by Mahmood et al [34], which demonstrated that patient-centered communication significantly reduces frustration during information-seeking among cancer survivors. It cannot be overlooked that while OHIS has to some extent challenged the traditional paternalistic medical model, its role in clinical practice is increasingly being reinterpreted. Far from threatening physician authority, it may actually enhance the depth and quality of doctor-patient communication [35]. Furthermore, patients’ heightened focus on dietary adjustments and nutritional therapies when evaluating information is deeply rooted in local health beliefs and traditional practices. This provides empirical support for moving beyond universal health literacy models and constructing culturally sensitive analytical frameworks.

In contrast to the results of the review by Chen et al [8], our study found that Chinese patients with breast cancer did not use the websites of national or nonprofit cancer organizations to search for health information. Chen et al [8] showed that patients with breast cancer, including those in China, often relied on these websites for information. This disparity may indicate that the patients in this sample have relatively low digital health literacy and have limited access to information through professional web-based channels. Digital health literacy specifically refers to a series of technical abilities and cognitive skills that individuals need to discover, understand, evaluate, and apply web-based health information [36]. The patients with breast cancer in this study’s reliance on social media and general search engines may directly indicate their deficiencies in the 2 key aspects of digital literacy, namely identifying the authority of information and conducting effective web-based navigation. In this context, digital health care approaches, such as hospital official apps, web-based patient communities, or intelligent question-answering systems, can provide patients with alternative channels for information acquisition, thereby reducing the obstacles caused by insufficient digital literacy [37,38]. The demand of patients with cancer for digital health functions is closely related to their digital health literacy levels, and there are also differences in the preferences for digital health interventions among patients with different literacy levels. This indicates that digital literacy is an important factor influencing the use of digital nursing tools [39]. Through these platforms, patients can not only access timely and reliable health information but also pose questions, verify information, and receive personalized advice in an interactive environment, which in practice contributes to enhancing their ability to understand and apply information. Meanwhile, research indicates that there are significant disparities in digital health literacy among breast cancer patients. Patients with low literacy levels face greater difficulties in participating in eHealth interventions, highlighting the need to consider patients’ literacy levels when designing digital nursing resources [40].

One of our findings is that the web-based information search behavior of patients with breast cancer changes with the disease stage, similar to the results of Madge, but their research focused on patients who had undergone breast cancer surgery [41]. Our participants had a wider range of disease stages, including both patients who had just undergone breast cancer surgery and those with recurrent breast cancer. The phased changes in information search behavior were more significant, suggesting that web-based health information intervention should be time-sensitive.

What is worth noting is that many participants with breast cancer, when searching for web-based health information, not only used traditional search engines and social media but also showed a high preference and trust in AI platforms. As shown in related studies, while traditional web-based resources remain dominant, LLM-based chatbots are becoming a resource for health information for some users, especially those who are younger and have a higher level of trust in AI [42]. However, this AI preference also comes with potential risks. The study found that the web-based health advice provided by AI models contains information misguidance and inaccurate information sources [43]. It can be seen that AI preference may trigger new cognitive trust traps. Therefore, we suggest that when promoting AI-assisted information services, it must be incorporated into the collaborative doctor-patient communication process, and through educational courses, web-based lectures, etc, improve patients’ digital health literacy, thereby enhancing the accuracy of information acquisition and decision-making quality for patients [44].

This study’s findings align with key processes in the stress-coping theory, thereby providing empirical support for understanding this theory in digital health information contexts. Concurrently, this study theorizes the complex web-based health information environment itself as a significant digital stressor, thereby expanding the application context of stress and coping theories. Findings indicate that for patients, this environment serves not only as a repository of resources for seeking support but may also constitute a persistent source of psychological stress due to information overload, contradictions, and uncertainties. In essence, an overwhelming or inconsistent volume of information can trigger anxiety and stress responses, potentially leading to avoidance behaviors [45]. Research findings also indicate that information avoidance and reduced retrieval frequency are not merely passive reactions, but rather contextualized emotion regulation strategies. This discovery complements traditional coping classifications’ understanding of avoidance strategies. Unlike the original theory’s focus on immediate coping, this study found that patients construct meaning around their illness through repeated comparison and integration of information from multiple sources, thereby adjusting their expectations for the future. This demonstrates the long-term and developmental nature of the coping process. It is worth noting that since this theory was not originally designed for digital health information scenarios, some new phenomena related to web-based information conflict and digital uncertainty lack direct conceptual counterparts within the theory. The findings of this study provide new perspectives on the applicability and extension of this theory within digital health information environments.

This study has the following limitations. Although most participants came from different cities and prefectures within Hunan Province of China, there are differences in language, economic development, and urban-rural structure within the province. However, the sample size of the qualitative research is limited, and it may still not cover the diversity of digital literacy and socioeconomic backgrounds. Participants were patients from a specialized cancer hospital, and these findings may not fully reflect the experiences or perspectives of health care providers in other medical systems or regions with differing technological infrastructure. Although semistructured in-depth interviews were adopted to obtain data, the respondents’ recollections might have a selective bias, and some emotional or behavioral experiences might not be accurately expressed or recalled. At the same time, qualitative research itself has a strong interpretative and subjective nature, and different analysts might form different understandings of the same material. This study was conducted during a treatment stage of patients with breast cancer and did not longitudinally track the web-based information search process and behaviors of patients with breast cancer. Future research could further focus on the differences in information coping patterns among people from different cultures and with different health literacy levels, and also examine how coping strategies evolve dynamically over the course of the treatment from a longitudinal perspective. This study analyzed the research results based on the theory of stress and coping. It provided theoretical support for understanding web-based information-seeking behavior. However, these theories may have limitations in terms of applicability in the specific cultural context of China. Future research could further validate these theories by integrating cultural adaptation theories.

This study reveals the complex experiences of patients with breast cancer in navigating web-based health information in the digital age. These findings not only deepen our understanding of patients’ health information behaviors but also provide significant insights for improving health policies, optimizing clinical practices, and guiding future research directions. The key points are as follows:

The web-based information-seeking experience of Chinese patients with breast cancer is a dynamic process involving motivation, cognition, and adaptation. Their search behavior stems from dual needs: emotional drive and problem-solving. Within a complex and conflicting information environment, they assess authenticity and applicability while constructing meaning. Through strategies such as information transformation, emotional regulation, support network building, and expectation adjustment, they achieve individual adaptation. Theoretically, this finding enriches the research framework of digital health literacy and information behavior and provides empirical evidence for understanding patients’ coping mechanisms in a complex information environment. In practice, this study suggests that clinical and nursing staff should pay attention to patients’ emotional needs, cognitive judgment abilities, and adaptation strategies when designing digital health interventions and information support strategies. They should provide multichannel, easy-to-understand, and actionable information resources and combine social support and emotional regulation guidance to help patients obtain, understand, and apply health information more effectively.