This study used a descriptive qualitative design to explore the OHIS experiences of patients with prostate cancer receiving ADT. This approach is particularly suitable for obtaining practice-oriented and experience-near accounts of participants’ perspectives in naturalistic clinical contexts [29]. The study followed the Standards for Reporting Qualitative Research (SRQR) guidelines (Checklist 1) [30].

The research team included nurses with clinical experience in urological oncology: 3 researchers, 2 research assistants, and 2 clinical experts. The lead researcher (SH) had 3 years of clinical experience in a urology outpatient clinic and conducted all interviews within that setting. This insider positionality facilitated rapport with participants but also risked shaping data interpretation through pre-existing clinical assumptions. To manage this influence, the research team maintained reflexive awareness throughout data collection and analysis by using reflexive journals after interviews, conducting regular team discussions to critically examine emerging interpretations, and comparing independent coding perspectives across researchers with different professional roles. All team members had prior training in qualitative research methods.

This study was conducted from July to October 2025 in a tertiary hospital in Zhejiang Province, China. The hospital performs approximately 3000 prostate biopsies each year, ensuring a sufficient patient population for the study.

Participants were recruited using purposive sampling to ensure variation in age, education level, treatment duration, and health literacy. Inclusion criteria were as follows: (1) diagnosed with prostate cancer, (2) receiving ADT for ≥3 months, (3) aged 18 years or older, (4) able to communicate in Mandarin, and (5) willing to participate and provide informed consent. Exclusion criteria were cognitive impairment, severe mental illness, or other serious diseases that hindered participation (such as concurrent other malignant tumors or severe complications).

Participants were recruited from the urology outpatient clinic and inpatient department of the study hospital. Potentially eligible participants were first identified by the lead researcher (SH) through the hospital’s electronic medical record system according to the inclusion criteria. After confirming eligibility with the treating urologists in the outpatient clinic or the responsible nurses in the inpatient department, patients were introduced to the study by their clinician or nurse. Those who expressed interest were subsequently approached by the lead researcher, who explained the study purpose and procedures, reconfirmed eligibility, and obtained written informed consent before the interview.

The number of participants was determined according to the data saturation principle [31]. Saturation was assessed iteratively throughout data collection and analysis. After each interview, emerging participant characteristics were compared with the developing persona framework to identify features that were insufficiently represented or not adequately captured by existing persona patterns, consistent with a process of negative case analysis. By the 18th interview, no substantively new characteristic patterns relevant to persona construction were identified, and the existing framework was considered sufficient to capture the range of OHISB observed across the sample. Two additional interviews were subsequently conducted to confirm the adequacy of the framework. As no new persona characteristics or participant patterns emerged, a final sample of 20 participants was considered sufficient to support the identification of 4 representative personas.

Following written informed consent, basic demographic and clinical information, including age, education level, and treatment duration, was collected. Interviews were conducted in a quiet, private meeting room, except for 3 participants who opted for telephone interviews, and were generally scheduled immediately after outpatient consultations or, for inpatients, before hospital discharge. Each interview was conducted by 2 researchers (SH and ZQ), with one leading the interview and the other taking field notes and observing nonverbal cues. Interviews lasted approximately 30 to 60 minutes. With participants’ permission, all interviews were audio-recorded and transcribed verbatim within 24 hours. During the interviews, nonverbal information such as expressions, tone, and gestures of the participants was recorded, and timely follow-up questions and clarifications were made based on the participants’ statements.

A semistructured interview guide was developed based on the study objectives and refined through discussion with 1 urology nursing expert and 2 urology clinicians. Two pilot interviews were conducted to assess the clarity and appropriateness of the interview questions. Data from the pilot interviews were not included in the final analysis. The final interview guide is provided in Multimedia Appendix 1.

This study used an inductive content analysis approach to analyze the interview data [32]. This approach was selected because no existing theoretical framework adequately captured the heterogeneity of OHISB among patients with prostate cancer receiving ADT. Inductive analysis allowed participant characteristics and behavioral patterns to emerge directly from the data and subsequently informed persona development. MAXQDA (version 2020; VERBI Software) was used for data management and coding.

The analysis proceeded through 3 iterative stages. Stage 1 is open coding. The unit of analysis was the meaning unit, defined as words, sentences, or paragraphs containing a coherent idea related to OHISB. Two researchers (SH and XL) independently read the transcripts multiple times to achieve immersion in the data. Meaning units were identified and assigned initial codes that reflected their underlying meaning. Stage 2 is the formation of subcategories. The 2 researchers met weekly to compare independently generated codes and discuss similarities and differences in interpretation. Through constant comparison, conceptually related codes were grouped into more abstract subcategories. Stage 3 is abstraction of higher-order categories. Subcategories were further compared and synthesized through iterative team discussions. Through abstraction and constant comparison, 4 higher-order categories were identified that captured recurring patterns in participants’ OHIS experiences. These categories subsequently served as the analytical domains for persona construction. The detailed codebook is provided in Multimedia Appendix 2.

Persona construction was informed by user-centered persona approaches commonly used in health informatics research, in which representative archetypes are developed through systematic comparison of shared behavioral characteristics [18,19]. After coding, participant characteristics identified across the 4 analytical categories were summarized into a participant characteristic matrix (Multimedia Appendix 3). Participants were then systematically compared across this matrix to identify recurring patterns of similarity and difference. Participants sharing similar combinations of motivations, information-seeking preferences, barriers, and support needs were iteratively grouped and compared. Through this iterative comparison process, 4 representative personas reflecting distinct patterns of OHISB were identified.

Each persona was named according to its dominant behavioral characteristics. Narrative persona descriptions integrated the defining motivations, preferences, barriers, and support needs associated with each persona type. Illustrative portrait images were AI-generated based on synthesized textual descriptions to support the visual communication of generalized persona characteristics rather than represent identifiable individuals.

To enhance trustworthiness and analytic rigor, several strategies were used throughout the study. Investigator triangulation was achieved through independent coding by 2 researchers, followed by regular comparison and discussion of interpretations. When disagreements could not be resolved through discussion between the 2 primary coders, a third senior researcher (WW) was consulted. The third researcher independently reviewed the relevant data segments and the codes in question, and a final decision was reached through team discussion until consensus was achieved. Participant validation and expert review were additionally used to strengthen the credibility and interpretive resonance of the findings.

Before participant validation, 2 clinical experts (an oncology head nurse and a urologist) reviewed the preliminary personas. They evaluated whether the personas reflected clinically recognizable patterns and whether the identified characteristics, barriers, and support needs aligned with their clinical observations. Based on their feedback, descriptions of digital literacy and information-seeking frequency were further clarified in Multimedia Appendix 3. Any discrepancies between expert feedback and the research team’s interpretations were resolved through joint re-examination of the relevant data and iterative discussion within the research team.

All 20 participants were invited to participate in member checking, and 17 responded. Validation occurred in 2 stages. First, preliminary analytical categories were shared with participants. Second, after persona construction, participants reviewed summary descriptions of the personas and were asked whether the personas reflected their experiences. Feedback was collected through outpatient conversations and WeChat. Fifteen participants considered the personas to meaningfully reflect their experiences, whereas 2 suggested minor revisions that were incorporated into the final descriptions (eg, incorporating descriptions of family dynamics, expanding emotional support needs, and softening negative characteristics in one persona type). This process constituted member checking of interpretations rather than factual verification of quotations, which had already been confirmed during the interviews.

This research was approved by the Clinical Research Ethics Committee of the First Affiliated Hospital of Zhejiang University School of Medicine (approval number 2025‐0202). All participants signed informed consent forms before participation. Each participant was assigned a unique code (eg, P1-P20) to ensure anonymity. Throughout the research process, the information and privacy of the participants were strictly confidential. All recordings were used for research purposes only and were not disclosed or used for other purposes without the consent of the participants. As compensation for their time and contribution, each participant received a small gift (valued at approximately RMB 20‐50) and was offered a brief disease-related consultation by the clinical team following the interview.

A total of 20 eligible patients receiving ADT participated in this study (Table 1). Participants ranged in age from 57 to 89 years, with a mean age of 70.5 (SD 9.4) years. The duration of ADT ranged from 3 to 65 months, with a mean duration of 21.4 (SD 19.1) months. Fourteen participants received ADT combined with radical prostatectomy, whereas 6 received ADT alone.

Based on recurring patterns across motivations, information access preferences, barriers, and support needs, 4 personas of OHISB were identified among patients receiving ADT. Table 2 presents an illustrative example of how participant characteristics were synthesized into a persona type, and Table 3 summarizes the defining characteristics of the 4 identified personas. Illustrative portraits were included to support visualization of the personas and do not represent actual participants.

These patients were mainly middle-aged and older men aged 57 to 73 years, who were relatively young prostate cancer survivors. They typically had a moderate or high level of education, good digital literacy, and tended to be proactive in disease management. Most of these patients were in the early-to-mid stages of the disease, emphasizing the use of information access to support decision-making and optimize communication with doctors. From the early stages of diagnosis, patients sought information about treatment options, medication differences, and side effect mechanisms while continuing to self-monitor.

Motivations: Their primary motivation was to reduce uncertainty and maintain a sense of control over the disease trajectory. They hoped to better understand treatment processes and participate more actively in decision-making and long-term management.

Barriers: Despite their active engagement, these patients frequently experienced information overload and fragmented content. They often had to compare information across multiple platforms and independently evaluate the credibility of different sources.

Needs: These patients preferred well-structured and authoritative online information platforms covering treatment processes, recurrence and metastasis, symptom management, medication differences, and long-term monitoring. They emphasized the importance of platform credibility and expressed a strong preference for hospital-recommended information sources and functions such as symptom tracking and prostate-specific antigen (PSA) monitoring.

These patients varied in age and educational background and generally demonstrated moderate digital literacy. They usually paid moderate attention to their health conditions and tended to avoid sensitive health information related to topics such as prognosis and changes in male characteristics. However, when sudden or worsening side effects of ADT (such as hot flashes and joint pain) affected their daily lives or their sense of male identity, it triggered anxiety in a short period of time, leading to an intensive and short-term online search state.

Motivations: Their information-seeking was mainly triggered by symptom-related uncertainty. They were eager to determine whether their symptoms were severe and how to relieve them quickly to restore a sense of stability in their daily lives.

Barriers: These patients were highly sensitive to frightening or conflicting online information, which often increased anxiety and led them to discontinue searching altogether. They also tended to avoid discussing sensitive issues, such as breast enlargement or sexual dysfunction, directly with health care professionals or family members.

Needs: These patients preferred concise, emotionally supportive, and easy-to-understand information presented through diagrams, short videos, or question-and-answer formats. The content should cover psychological support, side effect management, medication consultation, and interpretation of clinical indicators. Participants also valued anonymous peer communication and stepwise information delivery, which reduced the emotional burden.

Most of these patients have limited digital literacy and have received a shorter duration of ADT treatment. They have difficulties in understanding and identifying online medical information, so they tend to avoid directly processing complex treatment-related information. They usually have a relatively stable and highly participatory family support system, which can assume the functions of information management and decision-making assistance. Patients mainly play the role of treatment executors, while family members are responsible for screening, interpreting, and transmitting health information.

Motivations: Their main motivation was to reduce the cognitive and practical burden of independently managing complex medical information. Family involvement was viewed as a trusted and efficient way to support treatment management, thereby maintaining role harmony within the family.

Barriers: The main barriers included limited digital skills, difficulty in understanding online medical terminology, and concerns about misinformation or internet scams. Some participants also reported that information became incomplete or simplified during family-mediated communication.

Needs: These patients preferred family-inclusive information support systems that could synchronize reminders, medication schedules, and follow-up information across family members. They also valued clear visual instructions and accessible educational materials that could support understanding when family members were unavailable.

These patients generally had limited educational backgrounds, low digital literacy, and minimal active engagement with online health information. Most relied on doctors, family members, or occasional social media exposure for disease-related knowledge. They generally followed medical instructions but lacked a comprehensive understanding of ADT treatment and disease management.

Motivations: Their primary motivation was to maintain daily stability and avoid the burden associated with complex medical information. They tended to trust health care professionals and preferred following instructions rather than independently seeking information.

Barriers: These patients experienced substantial barriers related to low digital literacy, limited comprehension ability, and concerns about misinformation. Fatalistic beliefs about illness and aging further reduced their willingness to seek information.

Needs: They preferred low-threshold and easy-to-use health information support focused on practical daily management, including medication guidance, lifestyle advice, follow-up reminders, and reimbursement information. Participants also preferred large fonts, visual or audio materials, and continued access to offline support (face-to-face communication, paper materials). It is necessary to combine text messages and phone reminders to proactively push key information.

This study identified 4 personas of OHISB among patients with prostate cancer receiving ADT: proactive information managers, anxiety avoiders, family-mediated seekers, and passive or dependent seekers. Rather than representing fixed patient categories, these personas reflect recurring patterns of information engagement shaped by patients’ emotional responses, treatment experiences, sociocultural context, and perceived ability to manage health information. The findings illustrate the diversity of OHIS experiences during long-term ADT and provide a structured framework for understanding how informational, emotional, social, and contextual factors may interact to influence patients’ engagement with online health information. Recognizing these differing patterns may help inform the development of more responsive and patient-centered information support strategies for men receiving ADT.

Participants classified as proactive information managers demonstrated high levels of disease engagement and digital health literacy, consistent with previous quantitative studies indicating that younger age and higher education are commonly associated with greater information source use among patients with prostate cancer [14,33]. However, the present study further suggests that frequent information-seeking does not necessarily translate into effective information use. Highly engaged patients often assumed substantial responsibility for independently evaluating and integrating complex online information during long-term treatment management, creating a considerable cognitive burden [34]. These findings suggest that improving information organization, source transparency, and navigation support may be more beneficial for this group than simply increasing information availability [10,35]. In clinical practice, health care professionals may help reduce information overload by discussing patients’ online information use during routine follow-up consultations, recommending credible resources, and clarifying misunderstandings. Institutionally endorsed information platforms may also help facilitate patients’ navigation of complex online information during long-term treatment management [36,37].

Anxiety avoiders demonstrated a symptom-triggered and emotionally fluctuating pattern of information engagement, characterized by repeated shifts between intensive searching and abrupt withdrawal. This finding suggests that OHISB may function not only as a cognitive process for acquiring knowledge but also as a strategy for emotional coping under conditions of uncertainty. Previous studies have shown that ADT-related bodily changes, including gynecomastia and sexual dysfunction, may threaten masculine identity, self-image, and autonomy [38]. The present study extends these findings by illustrating how emotional vulnerability may shape patients’ interactions with online information environments. Some participants appeared particularly sensitive to frightening or conflicting information, which intensified uncertainty and contributed to avoidance behaviors [12]. This pattern also reflects the influence of masculine communication norms on OHISB. Consistent with previous evidence that men may avoid direct disclosure of vulnerability and emotional distress, some participants preferred anonymous online communities over discussing sensitive symptoms with clinicians or family members [39]. Within the Chinese cultural context, masculine norms emphasizing emotional restraint and endurance may further reinforce these tendencies [40], highlighting the importance of culturally sensitive information support. Previous studies have also shown that seeking anxiety may delay the collection of treatment information for prostate cancer survivors [41]. Information support for this group may therefore need to address emotional safety in addition to informational needs. Gradual information delivery, emotionally supportive communication, and proactive discussion of sensitive side effects by health care professionals may help patients maintain engagement while reducing distress associated with long-term ADT management [8].

The family-mediated seeker persona highlights the influence of family involvement and health care culture on OHISB within the Chinese context. Previous studies have shown that Chinese families often approach cancer as a collective experience, which may strengthen family cohesion while simultaneously shaping patients’ involvement in health-related decision-making [40]. The present findings suggest that, within family-centered care contexts, engagement with online health information may become a shared and negotiated process rather than an entirely individual activity. In some cases, patients’ access to online information appeared to depend substantially on how family members selected, interpreted, and communicated information during treatment management [42]. Structural characteristics of the Chinese health care system, including limited consultation time, high patient volume, and relatively doctor-centered communication patterns, may further reinforce reliance on family-mediated information support [43]. At the same time, family involvement should not be viewed solely as a barrier to patient autonomy. For many participants, family members represented important sources of emotional reassurance, practical assistance, and treatment coordination. Future digital information support systems may therefore benefit from family-inclusive approaches that engage caregivers while also preserving opportunities for patients to participate directly in the information process through shared platforms, synchronized reminders, and educational resources designed for both patients and caregivers [44].

Passive or dependent seekers in this study appeared to reflect not only limited digital capability but also a preference for minimizing emotional and cognitive burden during long-term illness management. In addition, some participants appeared to view illness progression and aging as largely unavoidable life processes, which reduced the perceived value of actively searching for additional health information. Previous research among Chinese older adults has similarly suggested that beliefs emphasizing acceptance of illness and limited personal control may be associated with lower motivation to seek health information [45]. Although such beliefs may reduce engagement with online resources, they may also represent a culturally and personally meaningful way of maintaining emotional equilibrium during long-term illness management. From this perspective, passive information–seeking behavior should not be interpreted simply as a lack of motivation or disinterest. Rather, for some participants, limiting information engagement appeared to function as an adaptive strategy for maintaining psychological stability in the context of overwhelming or difficult-to-interpret information environments [46]. These findings highlight the importance of respecting patients’ preferred level of information engagement rather than assuming that more information-seeking is always beneficial. Information support for this group may therefore need to prioritize accessibility, clarity, and low-burden communication [9]. Simplified medical language, multimedia educational materials, age-friendly digital interfaces, and sustained availability of offline support channels may help reduce barriers to understandable health information [13].

The findings of this study suggest that OHISB among patients with prostate cancer receiving ADT is shaped by psychological, sociocultural, and capability-related factors. The 4 personas identified in this study may provide implications for practice. In clinical settings, a brief assessment of patients’ preferred information engagement style, digital confidence, emotional responses to health information, and family involvement during routine follow-up consultations may help health care professionals identify patients who require different forms of information support. Rather than uniformly encouraging greater information-seeking, clinicians may need to tailor communication strategies according to patients’ preferred level of engagement, emotional readiness, digital capability, and individual beliefs about illness and personal control. The findings also highlight the importance of culturally sensitive and family-inclusive approaches to information support. Information interventions for patients receiving ADT may benefit from incorporating emotional support, proactive discussion of sensitive side effects, and opportunities for anonymous or low-pressure communication. At the same time, involving family caregivers in information support processes may improve continuity of care while preserving opportunities for patient participation and autonomy. At the digital health system level, future interventions may benefit from combining multiple modes of information delivery, including structured educational platforms, simplified multimedia materials, caregiver-linked systems, and offline support options.

This study has several limitations. First, all participants were recruited from a single tertiary hospital in China, which limits the geographic and institutional transferability of the identified personas. Differences in health care systems, digital health infrastructure, and cultural attitudes across regions may influence how the findings apply to other settings and populations. Second, the personas were constructed from self-reported interview data rather than objective records of actual digital behavior. While qualitative interviews provide the depth and contextual richness necessary for exploratory persona research, future studies combining interview data with platform usage records or digital trace data could offer a more complete account of patients’ information behavior. Third, the 4 personas represent a preliminary qualitative typology derived from a specific clinical context. Larger-scale quantitative and mixed methods studies across diverse populations are needed to further examine the prevalence, stability, and clinical applicability of these personas and to evaluate whether persona-informed information supports improving patient-centered outcomes.

This study identified 4 personas of OHISB among men with prostate cancer receiving ADT, demonstrating meaningful variation in patients’ motivations, information access preferences, barriers, and support needs. The findings highlight the influence of family-centered care practices and masculine communication norms within the Chinese sociocultural context on how patients engage with online health information. Rather than representing fixed patient categories, these personas provide exploratory and practice-oriented representations of recurring information engagement patterns during long-term ADT management. The findings suggest that future digital health information support may benefit from moving beyond uniform approaches and responding to patients’ diverse information engagement styles through more adaptive, culturally sensitive, and patient-centered strategies. However, further quantitative research is needed to examine the stability, generalizability, and clinical applicability of these personas across different health care settings and populations.