Barriers to home-based self-management in hospitalized patients with type 2 diabetes: a qualitative descriptive content analysis guided by protection motivation theory Authors: Zuyao Xiao, Ruiqing Cui, Shiyu Yang, Tao Liu, Ying Liu Published: 2026 ABSTRACT AbstractBackgroundEffective self-management is essential for glycemic control in type 2 diabetes. However, many patients experience self-management failure prior to hospitalization. This study aimed to explore the barriers to home-based self-management among hospitalized Chinese patients and interpret these barriers using Protection Motivation Theory (PMT).MethodsA descriptive qualitative research design was employed. From December 2025 to February 2026, semi-structured interviews were conducted with 19 patients with type 2 diabetes recruited through purposive sampling from a tertiary hospital in Northeast China. Data were analyzed using Graneheim and Lundman’s qualitative content analysis approach. The study was reported following the COREQ guidelines.ResultsSystematic content analysis of 19 interview transcripts yielded 319 initial codes derived from condensed meaning units identified through repeated reading of raw data. Following Graneheim and Lundman’s method, these codes were organized into 19 subcategories nested within 6 main categories at the manifest content level. Subsequent interpretive analysis, guided by Protection Motivation Theory (PMT), mapped these inductively derived categories to latent-level theoretical dimensions, elucidating the underlying psychosocial and cultural mechanisms driving self-management failure.ConclusionFailure in home-based self-management results from complex cognitive, psychosocial, and environmental barriers. Future nursing interventions should prioritize nurse-led transitional care models and culturally tailored health education to enhance patients’ self-efficacy and risk perception.Trial registrationNot applicable.Supplementary InformationThe online version contains supplementary material available at 10.1186/s12912-026-04873-9.Keywords: Type 2 diabetes, Qualitative research, Self-management, Qualitative description, Protection motivation theory Background Diabetes is a group of chronic metabolic diseases characterized by chronic hyperglycemia [1]. Diabetes not only poses a significant threat to individual health but also imposes a substantial socioeconomic burden. Relevant statistics indicate that both the number of people with diabetes and the associated healthcare costs in China will continue to increase [2–4]. In response to this challenge, China launched the National Diabetes Prevention and Control Action (DCPA) under the Healthy China 2030 initiative, promoting the standardization of diabetes management [5]. However, as a lifelong chronic condition, type 2 diabetes still requires long-term self-management to effectively control the disease [6]. Therefore, conducting an in-depth investigation into the barriers to self-management among people with diabetes prior to hospitalization is both urgent and crucial for alleviating the heavy burden of the disease. Diabetes self-management (DSM)—encompassing medication adherence, dietary control, physical activity, and blood glucose monitoring—is fundamental to glycemic control and prevention of complications in type 2 diabetes [6–8]. The American Association of Diabetes Educators (AADE) has established seven standard guidelines for DSM [9]. Although the beneficial effects of self-management behaviors in reducing complication rates are well established [10–12], many people with type 2 diabetes encounter substantial barriers to sustained adherence. Existing studies have identified factors such as healthcare system quality, social support, household income, educational attainment, and government investment as significant influences on self-management [13–21]. In China, the challenge is particularly acute: surveys indicate that self-management levels among type 2 diabetes patients remain significantly below optimal, with only a small proportion achieving sustained self-management behaviors [22, 23]. While these quantitative studies have identified multiple influencing factors, qualitative research exploring the psychological, social, and cultural mechanisms underlying self-management barriers in the Chinese context remains limited. Existing qualitative research has addressed selected dimensions—such as “face” (Mianzi) culture [24], family dynamics [25], diabetes stigma [26], and collectivist dietary norms [27]—but predominantly in community settings. Few qualitative studies have examined the pre-hospitalization home-care phase among hospitalized patients [22]. In-depth qualitative investigation of hospitalized patients’ experiences is needed to understand not merely what barriers exist, but why Chinese patients struggle with sustained self-management, particularly during the critical home-care phase preceding hospitalization. To address this knowledge gap, we conducted a descriptive qualitative study exploring barriers to home-based self-management among hospitalized Chinese patients with type 2 diabetes, with three specific objectives: (1) describe patients’ lived experience of self-management barriers during the pre-hospitalization home phase; (2) interpret these barriers through Protection Motivation Theory (PMT) to elucidate how inadequate threat and coping appraisals drive failure; and (3) illuminate psychosocial and cultural mechanisms—particularly the tension between individual health priorities and collectivist social obligations—shaping self-management decisions. We selected PMT because it distinguishes threat appraisal (severity, susceptibility, rewards) from coping appraisal (efficacy, response efficacy, costs) [28, 29], capturing both motivational and environmental barriers. While PMT has proven useful in diabetes prevention [28], its application in collectivist contexts remains limited. Unlike the Health Belief Model, PMT explicitly incorporates psychosocial costs of violating social norms—central to Chinese patients’ self-management decisions [29]. PMT served dual roles: during design, it informed the interview guide; during analysis, inductively derived categories were mapped to PMT dimensions. This approach balanced theoretical rigor with qualitative authenticity. Methods Research methodology This study employed a qualitative descriptive design using content analysis [30]. Descriptive qualitative research involves summarizing and describing the everyday discourse of respondents using systematic coding and categorization. We utilized the qualitative content analysis method developed by Graneheim and Lundman [31], which involves systematic analysis of textual data to identify and organize meaning into categories and subcategories. Methodological Positioning—Inductive-Deductive Mixed Strategy: This research adopted an explicitly mixed inductive-deductive analytical strategy, which warrants clear methodological positioning. During the initial coding phase, analysis was predominantly inductive, allowing participants’ lived experiences to emerge naturally without a priori imposition of theoretical constructs. Following the organic emergence of main categories through inductive analysis, Protection Motivation Theory (PMT) was applied deductively as an interpretive framework to map and synthesize these inductively derived categories (e.g., mapping ‘Insufficient risk perception’ to PMT’s Threat Appraisal dimension). This mixed approach is distinct from: (1) purely theory-driven content analysis, which typically imposes predetermined theoretical categories during initial coding, and (2) purely inductive descriptive analysis, which forgoes theoretical interpretation entirely. By sequencing inductive category generation before deductive PMT mapping, this approach preserves the authenticity of participant narratives while achieving theoretical coherence and depth of interpretation. Study Design Rationale: Adopting this design enabled us to collect direct, experience-near accounts from patients about barriers to self-management during the pre-hospitalization home care phase. PMT’s construct of threat appraisal (severity, susceptibility) and coping appraisal (efficacy, response efficacy, costs) provided a systematic framework for organizing and interpreting these complex, multifactorial barriers, particularly within the Chinese collectivist context, where social costs and normative pressures substantially influence health decision-making. Design of the interview guide The interview guide was developed by three nursing graduate students based on a literature review, using the key constructs of the PMT as a reference for the scope of questions to ensure that the interview fully explores the barriers to home self-management from the patient’s perspective. The guide was collaboratively drafted by the research team and reviewed by the supervisor to produce a preliminary draft. To avoid over-reliance on theory, the interview commenced with an open-ended question, allowing patients to freely describe the difficulties they had experienced with home-based management prior to hospitalization. Once the patient had finished describing their experiences, we then posed probing and clarifying questions based on the dimensions of the PMT framework, rather than requiring the patient to respond strictly according to the theoretical framework. In addition, we designed situational prompts (for example, days off, workdays, family dinners, dining out, purchasing medication, and blood glucose monitoring) to enhance the accuracy of patients’ recollections. At the end of the interview, additional open-ended questions were included to capture factors beyond the PMT framework, such as: “Are there any other difficulties we haven’t discussed?” and “What kind of support do you hope to receive from your family, the community, or the hospital in the future?” Prior to the formal interviews, we conducted pilot interviews (n = 3) to test the interview protocol, duration, and comprehensibility of the questions. Prior to the pilot interviews, we clearly informed patients that the study was unrelated to their treatment, that participation was voluntary, that they would be asked to sign an informed consent form, and that they could withdraw at any time. The pilot interviews were not included in the formal data set. After the pilot interviews concluded, the research team transcribed and reviewed the content to finalize the interview outline. (See Table 1) The final version of the interview outline was approved by 5 experts. The semi-structured interview guide was developed specifically for this study and has not been published previously. An English-language version of the guide is provided as Supplementary File 1. Table 1Interview guideSemi-structured Interview Guide1. What difficulties did you encounter while managing your blood glucose at home prior to hospitalization?2. Prompting questions (threat appraisal): What impacts or risks do you think diabetes poses? Are you worried about developing complications in the future? Why do you think so?3. Prompting questions (coping appraisal): What blood sugar management measures are you familiar with? Do you think these measures are effective? Are you confident that you can stick to your blood sugar control plan? In what areas do you think you might face difficulties?4. Prompting questions (cost-related): Do you feel that managing your blood sugar is a waste of time and energy? Do you consider the financial burden of diabetes to be significant? Do you find it more difficult to manage your blood sugar during family gatherings, dining out, or work-related social events?5. Are there any other challenges you didn’t mention during the interview? What kind of support would you like to receive from your family, society, or the hospital in the future? Interview guide Participant Recruitment: We selected the First Affiliated Hospital of Jinzhou Medical University for patient recruitment, with the recruitment period spanning December 2025 to February 2026. Using purposive sampling and maximum variation sampling (based on age, duration of diabetes, education level, and urban/rural residence), we identified patients with type 2 diabetes from the Geriatric Medicine and Endocrinology departments at the First Affiliated Hospital of Jinzhou Medical University through electronic medical records. Inclusion criteria were as follows: (1) diagnosed with type 2 diabetes according to the ADA’s ‘Standards of Care in Diabetes’ [32], (2) alert, fully oriented, and possessing full cognitive capacity to communicate effectively and make independent decisions. Exclusion criteria were: (1) experiencing acute critical illness or severe psychiatric disorders; and (2) unconsciousness or cognitive impairment (For example, dementia) that hindered independent communication.“(3) The participant showed any behavior that prevented the continuation of the interview, such as aggression. Efforts were made to select patients with diverse demographic characteristics. Participants were recruited on an ongoing basis. Transcriptions were completed within 24 hours of each interview. Initial coding was independently performed by two coders (Zuyao Xiao and Ruiqing Cui) on the transcripts. Preliminary analysis was conducted concurrently with data collection to enable ongoing assessment of sample adequacy. Sample size justification—information power framework Rather than invoking “data saturation”—which is incompatible with Graneheim and Lundman’s tradition—we employed Malterud et al.‘s (2016) information power framework [33]. This framework assesses sample adequacy across four dimensions: (1) study aim specificity; (2) sample specificity; (3) dialogue quality; and (4) analytical strategy. Our assessment is detailed below. Application of information power criteria Study Aim Specificity: Exploring barriers to self-management among hospitalized patients with T2DM is a moderately focused aim (neither overly narrow nor entirely open-ended) and requires a moderate sample size. Sample Specificity: Purposive sampling with maximum variation (age 21–80 years; disease duration < 1–>20 years; education from elementary to bachelor’s degree; urban and rural residence) yielded a moderate level of within-group homogeneity, appropriate for descriptive content analysis. Dialogue Quality: All 19 interviews were conducted by a single trained interviewer (ZYX) who established pre-interview rapport with participants and explicitly minimized power imbalance through clear assurances of research independence. Open-ended questions with reflective follow-up probing enabled deep engagement, yielding high-quality dialogue. Analytical Strategy: Graneheim and Lundman’s systematic coding, combined with investigator triangulation (two independent coders, ZYX and CRQ) and team-based consensus discussions, maximized analytical rigor and the efficiency of information extraction. Evidence of adequacy Given these factors, n = 19 was deemed to provide adequate information power. Confirmatory evidence: Initial coding (interviews 1–15) yielded stable category formation; no novel codes emerged after interview 15. Interviews 16–19 reinforced existing categories while revealing increased contextual depth (e.g., demographic moderation of barrier prominence). This pattern reflects information power sufficiency—a transition from “new code generation” to “nuanced contextual understanding”—rather than data saturation. Final sample Of 23 initially recruited patients, 4 declined participation prior to interviews, resulting in a final interviewed sample of n = 19. (Comprehensive qualitative rigor criteria justifying sample adequacy are provided in Sect. 2.6.) Data collection The interviews were conducted by the first author (Zuyao Xiao, ZYX; male, Master’s student in nursing), who received training in semi-structured interviewing prior to the formal interviews. The first author was not involved in the patients’ clinical treatment or care. Given the potential for power imbalances associated with conducting interviews in a hospital setting, we informed patients prior to the formal interviews that the study was independent of their clinical treatment, that participation was entirely voluntary, and that withdrawing or refusing to participate would not affect their treatment. The interview process utilized open-ended questions and clarifying follow-ups, avoiding leading questions and evaluative responses. Interviews were conducted in patients’ wards during rest periods. The interviewer had established rapport with participants beforehand. All interviews were audio-recorded using a mobile phone; duration ranged from 15 to 35 min (median: 22 min; mean: 23 min). While this timeframe is relatively brief, several methodological factors supported adequate depth. First, pre-interview rapport-building enabled rapid engagement without lengthy introductions. Second, the interview guide commenced with an open-ended question allowing participant-directed narrative, with subsequent probes highly targeted to minimize superfluous conversation. Third, hospitalized patients’ heightened awareness of their health trajectory facilitates reflection on complex experiences. Finally, preliminary concurrent analysis indicated stable category formation by interview 15; interviews 16–19 reinforced existing categories while revealing contextual depth, suggesting sufficient information power. We acknowledge that categories requiring sustained exploration—such as existential costs or learned helplessness—may have benefited from extended time. Future research on trauma-related barriers should employ 45–60 min interviews. Translation All interviews were conducted in Mandarin Chinese. Following transcription, verbatim Chinese transcripts were independently translated into English by two bilingual researchers (both native Chinese speakers with fluent English proficiency). For culturally specific phrases or concepts that carry particular significance in Chinese but lack direct English equivalents (e.g., “面子” (Mianzi)—“face”), the original Chinese term was retained with a parenthetical English translation. Discrepancies between the two translations were resolved through discussion, prioritizing semantic and conceptual accuracy over literal word-for-word translation. The final translated transcripts were spot-checked against the original Chinese by a third bilingual team member to ensure fidelity. Critically, initial coding was performed on the original Chinese transcripts by the primary coder (ZYX) to preserve semantic nuance and cultural context. The secondary coder (CRQ) independently coded the English translations. When coding divergences emerged between the two coders, the research team examined the original Chinese text to determine whether differences reflected genuine analytical disagreement or translation artifacts. This bilingual parallel coding strategy ensured that culturally specific meanings and implicit linguistic relationships were retained rather than flattened through translation into generic English categories. For example, the Chinese concept of “没面子” (mei mianzi—literally “no face”) carries embedded meanings of honor, social standing, and collectivist obligation that risk oversimplification if coded solely as the English translations “shame” or “embarrassment.” By anchoring initial coding in the original Chinese while using English translations as a validity check, the analysis preserved both analytical rigor and cultural authenticity. Researcher characteristics and reflexivity To ensure qualitative rigor, the research team maintained continuous reflexivity throughout the study. All semi-structured interviews were conducted by the first author (Zuyao Xiao), a nursing graduate student trained extensively in qualitative research methods and interviewing techniques. Importantly, the interviewer was independent of the clinical care team providing direct treatment to the participants. We recognized the potential for an inherent power imbalance and social desirability bias—specifically, the tendency for hospitalized patients to underreport their self-management failures when speaking to a healthcare professional. To mitigate this, the interviewer spent time on the ward prior to data collection to establish rapport with the patients. At the start of each interview, it was explicitly emphasized that the dialogue was for research purposes, highly confidential, and detached from their clinical care. The interviewer adopted a patient, empathetic, and non-judgmental stance, using open-ended questions to encourage honest reflections. Concrete Illustration of Stance Adjustment: A practical example illustrates how nursing background bias was actively managed. During the interview with Participant P, the first author (ZYX) initially interpreted the participant’s statement—“I feel that lifestyle habits are not very effective; only insulin can control it properly”—as a simple “knowledge deficit.” This interpretation reflected ZYX’s clinical nursing training, which prioritizes the effectiveness of lifestyle interventions alongside pharmacotherapy. However, upon reflection and team discussion, the research team recognized that the participant’s statement actually articulated a deeper psychological mechanism: learned helplessness stemming from previous failed attempts at dietary control. This reframing required conscious deprioritization of clinical assumptions and conscious re-positioning to honor the participant’s authentic lived experience rather than the interviewer’s professional framework. This reflexive process demonstrates how explicit awareness and team-based discussion prevented the researcher’s nursing background from reducing participant complexity to clinical categories. Furthermore, during the data analysis phase, the research team was conscious of their nursing backgrounds and prior knowledge of Protection Motivation Theory (PMT). To prevent clinical or theoretical assumptions from overshadowing the patients’ genuine voices, investigator triangulation was employed. Initial coding was performed inductively by two independent researchers, and theoretical mapping to PMT occurred only after the core categories had organically emerged from the data. Disagreements were resolved through reflective discussions with the corresponding author to ensure the interpretations remained strictly grounded in the participants’ lived experiences. Data analysis Within 24 hours of each interview’s completion, the audio recordings were transcribed, and the researcher listened to them to verify and correct the transcripts. This study conducted preliminary data processing concurrently with the interviews to support fine-tuning of the interview guide and assessment of sample adequacy. Upon completion of transcription, the data were imported into NVivo 15 for initial coding. This study employed the qualitative content analysis method developed by Graneheim and Lundman [31] following these steps: (1) repeatedly reading and becoming familiar with the raw data; (2) identifying meaning units relevant to the study; (3) condensing these meaning units to form initial codes; and (4) categorizing the initial codes to develop categories. Analytical Transparency—Exemplification of Graneheim and Lundman’s Method: To enhance transparency regarding the analytical process, Table 2 (below) provides a concrete illustration of how raw data were systematically transformed through each analytical step: This exemplification demonstrates how raw participant narratives were distilled into condensed meaning units that retain semantic fidelity, then encoded into theoretically neutral codes, and finally organized into analytically meaningful subcategories consistent with Graneheim and Lundman’s framework. To enhance dependability, initial coding was performed independently by two researchers: Zuyao Xiao (ZYX; male, first author, 18 months of clinical experience) as the primary coder, and Ruiqing Cui (CRQ; female, 16 months of clinical experience) as the second coder. Coding disagreements were resolved through structured team discussion to achieve consensus. Following the emergence of main categories through inductive analysis, the research team—including Shiyu Yang (SYY; female, 14 months of clinical experience)—participated in team discussions to systematically map inductively derived categories to Protection Motivation Theory (PMT) dimensions (e.g., mapping ‘Insufficient risk perception’ to Threat Appraisal: Perceived Severity/Vulnerability). SYY participated in this interpretive phase but not in initial coding. This dual-layered strategy—inductive category generation followed by deductive PMT mapping—ensured analytical rigor while achieving theoretical coherence. The final three-level coding system (codes → subcategories → main categories) was established through collective discussion among all three graduate students (ZYX, CRQ, SYY) and supervisors (TL, YL), and subsequently reviewed and confirmed by the second corresponding author (YL). Analysis was conducted using NVivo 15 software and guided by COREQ criteria to enhance transparency [34]. (See Appendix 1 for details.) Table 2Exemplification of the qualitative content analysis process according to Graneheim and LundmanRaw dataCondensed meaning unitCodeSubcategory“Unless it is very serious, I don’t go to the hospital for checkups. Plus, with my busy work schedule, I don’t think about these things.” (Interviewee J)The participant avoids clinical checkups unless experiencing severe symptoms, citing a busy work schedule as the primary reason for neglecting management.Neglecting checkups due to work scheduleWork-related stress“I don’t really drink with my family, but I have to drink with my friends. Otherwise, how can we keep in touch?” (Interviewee N)The participant consumes alcohol with friends to maintain social connections, despite not drinking at home, prioritizing social bonds over dietary restrictions.Drinking alcohol to maintain social relationshipsLack of social understanding and care“I don’t think dietary control is effective. No matter what I eat, my blood sugar stays around 10. I feel that controlling blood sugar still depends on taking medication and injecting insulin.” (Interviewee R)The participant believes dietary control fails to improve blood glucose levels, concluding that glycemic control is exclusively dependent on pharmacological treatment.Skepticism toward dietary control effectivenessDoubts about the effectiveness of self-management behaviors“You have to remember to inject insulin on time, be careful about what you eat, and sometimes you have to think about exercising—it’s a huge waste of time and energy.” (Interviewee I)The participant perceives the daily requirements of insulin injections, dietary vigilance, and exercise planning as an exhausting drain on personal resources.Management regimen perceived as exhaustingThe time and energy investmentNote: Each row represents one meaning unit derived from the interview transcripts of actual participants in this study. Multiple subcategories were subsequently aggregated into overarching main categories Exemplification of the qualitative content analysis process according to Graneheim and Lundman Note: Each row represents one meaning unit derived from the interview transcripts of actual participants in this study. Multiple subcategories were subsequently aggregated into overarching main categories Clarification on Manifest vs. Latent Analysis in This Study: Table 2 illustrates how raw data proceed through Graneheim and Lundman’s manifest-level coding steps (meaning units → condensed units → codes → subcategories). The examples shown—“face concerns,” “family control,” “collectivist dietary norms,” “diabetes stigma”—represent MANIFEST categories reflecting visible behavioral and psychological content. In the Discussion Sect. (4.1–4.6), these manifest categories are REINTERPRETED at the latent level through PMT, thereby answering deeper “how” and “why” questions: For instance, the manifest category “Lack of social understanding and care” (Sect. 3.23) becomes a latent-level theme in Discussion 4.2: the tension between Chinese collectivist values (face, harmony, obligation) and individual health imperatives—a psychological and cultural mechanism not explicitly named by participants but revealed through theoretical synthesis. This two-stage analytical movement—from manifest description to latent interpretation—is essential to Graneheim and Lundman’s framework when applied to exploratory research aimed at understanding complex phenomena. Rigor To ensure the rigor of this qualitative study, we adopted the criteria established by Guba and Lincoln [35], which include Credibility, Dependability, Transferability, and Confirmability. First, regarding the Credibility of the study, we ensured that open-ended questions and clarifying follow-ups were used during interviews; we did not interrupt patients or lead them, thereby maximizing the reflection of their genuine feelings and ensuring Credibility. Second, regarding Dependability, we conducted continuous and in-depth analysis of the interview data through independent coding by two researchers and repeatedly compared the codes. Additionally, the coders were proficient in using NVivo 15 software to enhance the stability and reproducibility of the research process. Third, regarding Transferability, we ensured this by recruiting patients with diverse demographic characteristics and establishing strict inclusion and exclusion criteria. Finally, regarding Confirmability, we ensured this by having the same researcher conduct both the preliminary and formal interviews and by establishing rapport with the patients prior to the interviews. This facilitated the interview process. Additionally, the interview results were discussed and finalized collectively by the research team, ensuring Confirmability. Member Checking: Member checking was not conducted in this study, as participants were discharged during preliminary analysis to geographically dispersed locations, and collecting post-discharge contact information would have imposed undue burden on vulnerable hospitalized patients. Instead, credibility was strengthened through investigator triangulation (independent coding by two researchers), prolonged engagement with data through repeated coding cycles, peer debriefing and reflexive discussion within the research team, and transparent documentation of analytical decisions. Field Notes: No written field notes were recorded during interviews, as maintaining continuous engagement with participants took priority. We acknowledge this as a limitation, as field notes can capture non-verbal cues, environmental context, and immediate researcher reflections that audio recordings alone may not preserve. To partially compensate, the interviewer conducted brief verbal reflective debriefs immediately following each interview, which were transcribed and reviewed during analysis. Additionally, audio recordings were reviewed multiple times during transcription, allowing coders to attend to paralinguistic markers such as hesitation, emphasis, and emotional tone. While data saturation is not applicable to Graneheim and Lundman’s approach, our concurrent preliminary analysis enabled continuous assessment of whether emerging categories possessed sufficient information power for rigorous interpretation. By interview 15, the categorical structure had crystallized (19 codes condensed into 19 subcategories nested in 6 main categories), and interviews 16–19 demonstrated stable reinforcement of these categories while revealing increased depth regarding demographic moderation of barriers (e.g., work-related stress disproportionately affecting working-age participants; existential costs more prominent in older retired participants). This pattern indicated sufficient information power for meaningful interpretation within our analytical framework. Ethical considerations This study was approved by the Ethics Committee of Jinzhou Medical University (Ethics Approval Number: JZMULL2026004) and strictly complied with the Declaration of Helsinki. Data collection took place from December 2025 to February 2026. Prior to the formal interviews, the researcher explicitly explained the study’s purpose, procedures, and confidentiality measures to the participants. It is important to clarify that all selected participants possessed full cognitive and decision-making capacity. Therefore, written informed consent was provided independently and solely by the patients themselves. While family members were often present and kept informed about the study—reflecting the strong family-centered care norms prevalent in Chinese hospital culture—family consent was not a legal or ethical requirement for participation. Participants were fully informed that their participation was entirely voluntary, and they could withdraw from the study at any time without any negative impact on their clinical treatment or care. Patient privacy was strictly protected, and no identifiable personal information is included in this manuscript. Results A total of 19 patients participated in this study. Participants’ ages ranged from 21 to 80 years, with diabetes duration spanning from newly diagnosed to 30 years (mean disease duration: 9.3 years). The demographic characteristics of the participants are summarized in Table 3. Demographic Distribution and Data Anchoring: Table 3 provides the demographic characteristics of the 19 participants. In the results sections below, participant identifiers are systematically linked to demographic data to enable readers to contextualize specific barriers and experiences. For example, mental health issues and work-related stress were particularly pronounced among younger and middle-aged working participants (ages 41–60 with active employment), while existential costs of management and concerns about complications were more prominent among older retired participants (ages 71–80 with longer disease duration). This demographic anchoring enhances the specificity of qualitative findings and permits readers to assess transferability of findings to their own populations. Table 3Demographic characteristics of participants (N = 19)CharacteristicCategoryn (%)Age range (years)21–301 (5.3)31–402 (10.5)41–501 (5.3)51–608 (42.1)61–703 (15.8)71–804 (21.1)Duration of illness< 1 year4 (21.1)1–5 years3 (15.8)6–10 years6 (31.6)11–20 years4 (21.1)> 20 years2 (10.5)OccupationPhysical labor7(36.8)Retired6 (31.6)Unemployed2 (10.5)Non-physical labor3(15.8)Self-employed1 (5.3)Marital statusMarried16 (84.2)Other3 (15.8)ResidenceUrban13 (68.4)Rural6 (31.6)EducationElementary school3 (15.8)Junior high school8 (42.1)High school4 (21.1)Associate degree3 (15.8)Bachelor’s degree1 (5.3)Other: Divorced, widowed, unmarried Demographic characteristics of participants (N = 19) Other: Divorced, widowed, unmarried Content analysis of the 19 interview transcripts generated 319 initial codes, which were systematically organized into 19 subcategories nested within 6 main categories, representing the primary barriers to home-based self-management among hospitalized Chinese patients with type 2 diabetes: (1) multifaceted stressors in the living environment, (2) gaps in the social support system, (3) insufficient perception of diabetes risk, (4) lack of knowledge about diabetes, (5) low self-efficacy, and (6) high costs of management. These findings are summarized in Table 4, which presents the categorical structure and number of codes within each subcategory. The detailed coding structure and mapping of the 319 initial codes to their corresponding subcategories are provided in Appendix 3 (Coding Tree). Additionally, the categories were mapped to the PMT framework and interpreted (see Appendix 2 for details). Detailed descriptions and illustrative participant accounts for each category are provided in the sections below. Table 4Categories and sub-categories related to barriers to home-based self-managementMain CategorySubcategoryNumber of codes (n = 319)Multifaceted stressors in the living environment.Lack of rests3Work-related stress12Mental health issues29Gaps in the social support systemInsufficient support from hospitals19Insufficient social support29Lack of social understanding and care12Insufficient perception of diabetes riskHerd mentality8Lack of warning signs regarding future complications9Insufficient awareness of the risks associated with diabetes26Lack of knowledge about diabetesLack of health literacy21Lack of knowledge about the risks of diabetes17Difficulty in distinguishing between accurate and inaccurate information online11Low self-efficacy.Lack of good self-management habits28Lack of confidence in self-management25Doubts about the effectiveness of self-management behaviors3Difficulty stepping outside one’s comfort zone12High costs of management.The time and energy investment18The heavy financial burden13The cost in terms of life’s meaning5Note: n represents the number of initial codes assigned to each subcategory. In qualitative content analysis, code frequency demonstrates systematic data distribution but does not solely determine category importance; theoretical and substantive significance are equally important criteria for category retention and interpretation Categories and sub-categories related to barriers to home-based self-management Note: n represents the number of initial codes assigned to each subcategory. In qualitative content analysis, code frequency demonstrates systematic data distribution but does not solely determine category importance; theoretical and substantive significance are equally important criteria for category retention and interpretation Note on Retention of Low-Frequency Categories: In qualitative research, category importance is determined by substantive and theoretical significance rather than frequency alone. Consequently, subcategories representing fewer codes (e.g., ‘Doubts about the effectiveness of self-management behaviors,’ n = 3) were retained when they revealed critical psychological mechanisms such as learned helplessness, providing essential insight into patients’ complete disengagement from self-management despite awareness of risks. Multifaceted stressors in the living environment This topic addresses the pressures patients face in their daily lives—such as those related to sleep, work, and emotions—which undermine their ability to manage their own care. Lack of rest Chronic sleep insufficiency, driven by both physiological tendencies (e.g., being a light sleeper) and noisy living environments, was highlighted as a disruptive factor for blood glucose control. For instance, Interviewee Q vividly illustrated environmental interference: “I don’t get much rest because my family runs a mahjong parlor.” Work-related stress Occupational demands severely constrained patients’ time and energy for daily self-management. For those engaged in physically demanding labor, maintaining work efficiency and securing income often superseded dietary control. As Interviewee O explained: “I can’t control my diabetes. My job is physically demanding, and if I don’t consume enough calories, I won’t be able to work, so I have no choice but to eat.” Mental health issues Pervasive psychological burdens—such as anxiety post-diagnosis, irritability, and a profound sense of helplessness—created an emotional tax that depleted patients’ motivation. Notably, many participants expressed deep guilt over the inconvenience their specialized dietary needs caused their families, which paradoxically hindered their commitment to self-management. Participant J (female, 51–60 years, 1–5 years disease duration, urban residence, self-employed) stated: “I still have to watch what I eat; my family cooks meals just for me… But doing this causes a lot of trouble for others.” Similarly, Participant I (male, 61–70 years, 11–20 years disease duration, urban residence, retired) articulated the emotional toll: “After being diagnosed with diabetes, I felt very anxious. I kept worrying whether my condition would worsen or if I would develop complications. This worry has been with me for years.” Gaps in the social support system This topic focuses on social issues, including insufficient hospital support, inadequate basic social assistance, and a lack of social understanding and care. These factors collectively hinder patients’ ability to manage their own care. Insufficient support from hospitals Patients highlighted a lack of comprehensive institutional support, citing inadequate post-discharge health education and negative treatment experiences as significant barriers. Unaddressed concerns about insulin-induced weight gain and inconsistent nursing skills diminished their adherence to treatment. As Interviewee C pointed out regarding injection techniques: “I’m afraid of pain. When I get insulin, some nurses have good technique and it doesn’t feel painful, but with others it feels really painful.” Insufficient social support Beyond the hospital setting, patients faced systemic logistical and financial barriers. Limited financial subsidies for long-term medications, a lack of community-based educational programs, and the geographic inconvenience of purchasing drugs collectively severely hindered their home-based care. Interviewee C emphasized the logistical burden: “It’s not very convenient to buy medication. Our home is quite far from the hospital, so coming here every time is a hassle.” Lack of social understanding and care Social norms and the cultural expectation of communal dining often conflicted with strict self-management. Adhering to dietary restrictions or refusing alcohol during gatherings was frequently perceived as unsociable. Consequently, patients routinely compromised their health disciplines to fulfill social obligations and avoid the stigma of isolation. Participant F (female, 51–60 years, 6–10 years disease duration, urban residence, retired) vividly expressed the psychological discomfort of maintaining boundaries at home: “When there are guests at home, or during holidays, I no longer cook separately for myself. I just eat with everyone else. During those times when I cooked and ate alone, it felt off psychologically.” This sentiment was echoed by Participant H (female, 51–60 years, 6–10 years disease duration, rural residence, physical laborer): “At work, colleagues often invite me out to eat. If I refuse, they say I’m being unfriendly or stuck-up. But when I go, I can’t stick to my diet. It’s difficult either way.” Insufficient perception of diabetes risk This topic highlights patients’ insufficient awareness of the long-term risks of diabetes, including herd mentality, a lack of early warning signs for future complications, and insufficient vigilance regarding the dangers of diabetes. Herd mentality Participants exhibited a pronounced herd mentality, normalizing unhealthy behaviors through peer imitation. By observing others engage in non-compliance without immediate repercussions, patients justified their own deviations from medical advice. Interviewee Q provided a vivid example of “gaming” the treatment regimen for social reasons: “For example, let’s say I’m going out for drinks tonight… I inject a few extra units of insulin before I start drinking. Then I go out—everyone does it this way.” Lack of warning signs regarding future complications A reactive rather than proactive mindset toward complications was prevalent. Some participants operated under an optimism bias, believing that complications were either avoidable through minimal effort or manageable only after they manifested. As Interviewee C succinctly expressed this fatalistic view: “I don’t know what complications there are; I’ll just treat them if they happen.” Insufficient awareness of the risks associated with diabetes A profound underestimation of diabetes-related risks was common, with many perceiving the condition as inherently “mild.” This insensitivity was frequently exacerbated when diabetes symptoms were masked or overshadowed by more overt comorbidities. For instance, Interviewee E noted: “Because I’ve had a stroke, many of the symptoms and complications of diabetes have been masked. I don’t feel that diabetes poses any harm either.” Lack of knowledge about diabetes This topic reflects resistance at the patient’s cognitive level. This includes insufficient health literacy, a lack of knowledge about the risks of diabetes, and difficulty distinguishing between accurate and inaccurate information online. Lack of health literacy Many patients exhibited a significant deficit in basic medical knowledge, often reducing the complexity of chronic disease management to oversimplified slogans. Beyond the vague concept of “eating less and exercising more,” they possessed few practical self-management skills. As Interviewee E summarized: “I basically know nothing about medical matters.” Lack of knowledge about the risks of diabetes A critical lack of awareness regarding the diverse and multifaceted complications of diabetes was evident. Patients typically relied on their immediate physical sensations (e.g., thirst or fatigue) to judge their condition, failing to recognize systemic risks that had not yet manifested as overt symptoms. Interviewee O explained: “I’ve heard from others that the risks of diabetes are very serious, but I really don’t know what they are specifically.” Difficulty in distinguishing between accurate and inaccurate information online While patients utilized digital platforms (e.g., smartphones and television) for information, they struggled with a profound sense of information distrust. The inability to distinguish credible medical advice from misinformation led to skepticism and even total disengagement from online health resources. Participant H (female, 51–60 years, 6–10 years disease duration, rural residence, physical laborer) noted: “I feel that a lot of the information online is helpful, but I remain skeptical about some of it because I suspect it might be false.” This information confusion was further illustrated by Participant O (male, 51–60 years, 11–20 years disease duration, urban residence, physical laborer): “I watch a lot of health videos on my phone. Some say eat this, some say don’t eat that. The advice is all contradictory. I don’t know who to believe, so I just ignore most of it.” Low self-efficacy This topic analyzes the factors that hinder patients’ self-management abilities from the perspective of self-efficacy. These include a lack of good self-management habits, insufficient confidence in self-management, doubts about the effectiveness of self-management, and difficulty stepping outside one’s comfort zone. Lack of good self-management habits Many participants exhibited persistent avoidance behaviors toward routine monitoring and clinical checkups. This was often driven by a combination of low health awareness and “ostrich-style” fear of receiving unfavorable results. Interviewee Q admitted: “I don’t come in for checkups because I’m afraid they’ll find some disease, so I just don’t come at all.” Lack of confidence in self-management Patients expressed significant anxiety regarding their ability to maintain self-discipline post-discharge without the structured supervision of a hospital environment. For many, the transition to home-based care felt overwhelming and unsustainable. Participant L (male, 51–60 years, 1–5 years disease duration, urban residence, non-physical laborer) explained: “Once I’m home, I certainly won’t have regular reminders to inject insulin or check my blood sugar like I do in the hospital.” This anxiety extended to the broader management regimen, as articulated by Participant E (male, 61–70 years, 6–10 years disease duration, rural residence, elementary school education): “The hospital arranges everything for me—the nurses remind me when to eat, when to take medication. At home, I have to manage everything myself. I’m not sure I can do it reliably.” Doubts about the effectiveness of self-management behaviors A critical barrier to engagement was a skepticism toward lifestyle interventions, often rooted in “learned helplessness.” Some patients believed that glycemic control depended solely on pharmacological interventions, viewing dietary or exercise efforts as futile. Interviewee P noted: “I feel that lifestyle habits are not very effective… only insulin can control it properly. I saw an experiment where people ate exactly the same thing, but their blood sugar was still different.” Difficulty stepping outside one’s comfort zone Long-standing dietary and lifestyle habits formed over decades presented a significant “comfort zone” that proved resistant to change. The perceived psychological cost of abandoning these life-long preferences led to persistent non-compliance. Interviewee R illustrated this entrenchment: “I love fruit. I can do without meals, but I have to eat fruit. I eat fruit instead of meals.” High costs of management This topic explores the costs of self-management in diabetes, including the time and energy investment, the heavy financial burden, and the cost to one’s meaning of life. The time and energy investment The majority of participants viewed diabetes management as an exhaustive daily burden. The constant cognitive and physical labor required for insulin synchronization, medication adherence, and dietary planning was perceived as a massive drain on their limited time and energy. Participant I (male, 61–70 years, 11–20 years disease duration, urban residence, retired) noted: “You have to remember to inject insulin on time, be careful about what you eat… it’s a huge waste of time and energy.” The cumulative fatigue was echoed by Participant O (male, 51–60 years, 11–20 years disease duration, urban residence, physical laborer): “After a full day of work, I’m exhausted. Then I come home and have to plan meals, check my blood sugar, figure out insulin doses. By the time I finish all that, I’m mentally drained. There’s no time left for anything else.” The heavy financial burden The cumulative cost of long-term pharmacological treatment, blood glucose monitoring supplies, and recurring clinical examinations imposed a severe financial strain. Participants expressed significant anxiety over these continuous expenditures, particularly when anticipating the added costs of future complications. Interviewee H highlighted the immediate burden: “It costs a lot. The tests and lab work are so expensive. I spent over 3,000 yuan just for a half-day hospital stay.” The cost in terms of life’s meaning Notably, some patients exhibited a form of existential pessimism, questioning whether the health benefits of strict self-discipline were worth the “cost” of losing life’s pleasures. For these individuals, the deprivation of dietary enjoyment was perceived as a loss of existence’s primary meaning. Participant R (male, 71–80 years, > 20 years disease duration, urban residence, retired) epitomized this trade-off: “It would be such a shame if I couldn’t eat what I want before I die. Everyone has to die sooner or later.” This sentiment was also articulated by Participant J (female, 51–60 years, 1–5 years disease duration, urban residence, self-employed), who framed the dilemma differently: “I’ve been strict with my diet for years, but I don’t feel like my quality of life has improved. I’m constantly worried and restricted. What’s the point of living a long life if I can’t enjoy it?” Discussion This study employed a descriptive qualitative methodology guided by Protection Motivation Theory (PMT) to explore barriers to home-based self-management among hospitalized Chinese patients with type 2 diabetes. Semi-structured interviews with 19 participants yielded 319 coded meaning units organized into six interconnected categories. The analysis revealed a dual deficit: inadequate threat appraisal (manifested as insufficient risk perception and knowledge deficits) coupled with compromised coping appraisal (manifested as low self-efficacy and high perceived costs). External environmental stressors and fragmented social support systems further constrained autonomous self-management. Notably, internal and external rewards did not emerge as discrete categories; instead, reward dimensions remained deeply embedded within social conformity pressures—particularly manifested through patients’ prioritization of face-saving and social harmony over individual health imperatives. This pattern suggests that when applying PMT in collectivist cultural contexts, the operationalization of “rewards” requires careful cultural adaptation. Sect. “Theoretical contributions: refining PMT in a collectivist context” elaborates on this theoretical refinement. The following Sect. (“Multifaceted stressors in the living environment”–“High costs of management”) systematically interpret each barrier category through PMT constructs and the Chinese sociocultural context. Multifaceted stressors in the living environment As one participant vividly articulated: “I don’t get much rest because my family runs a mahjong parlor” (Participant Q; female, 51–60 years, 6–10 years disease duration, urban residence, physical laborer), illustrating how environmental stressors directly fragment patients’ ability to engage in consistent self-management. Sleep deprivation, work-related stress, and psychological burdens collectively undermine patients’ capacity for health management. Previous studies have demonstrated that sleep deprivation leads to fluctuations in blood glucose levels [36], work-related stress increases the risk of developing diabetes [37], and negative emotions play a mediating role in patients’ self-management [38]. In this study, patients temporarily neglected diabetes management to “ensure work efficiency” and “maintain household income,” while long-term psychological burdens required them to spend significant time coping with negative emotions. These findings suggest that nurses should conduct a feasibility assessment rather than prescribe standardized regimens. Example: For a laborer who cannot reduce caloric intake: ‘Let’s focus on meal timing and glycemic index within a caloric target that sustains work performance. Existing studies have found that interventions such as guided progressive muscle relaxation [39], breathing rhythm regulation [40], and mindfulness practice [41] can gradually help them reduce stress in their daily lives. Gaps in the social support system One participant’s statement captures this tension: “When there are guests at home, or during holidays, I no longer cook separately for myself. I just eat with everyone else. During those times when I cooked and ate alone, it felt off psychologically,” (Participant F; female, 51–60 years, 6–10 years disease duration, urban residence, retired) revealing how social support gaps intersect with cultural expectations. Our data identified three distinct social barriers that impede patients’ self-management. Barrier 1, when hospital education is inadequate, or management is problematic, one interviewee stated: “I haven’t seen anyone offer me any advice at the hospital so far.” (Participant C; female, 61–70 years, > 20 years disease duration, rural residence, elementary school education) Prior studies on diabetes self-management education and support have indicated that systematic health education is crucial for improving patients’ self-management levels [7] and that the quality of clinician–patient communication can improve treatment outcomes [42]. Barrier 2: reduced access to social resources burdens patients with the responsibility of self-management, thereby diminishing their motivation to manage the disease. This corresponds with findings from other countries regarding the impact of public resources and government support on diabetes management [43, 44]. Barrier 3, this study found that patients lack understanding and support in social contexts, often sacrificing their diabetes self-management to maintain social decorum or avoid causing embarrassment. A deeper psychological analysis suggests this may be related to the conflict between Asia’s “culture of face” [24] and the “stigma of diabetes” [45]. Additionally, research by Ciarma and Mathew found that self-esteem mediates the relationship between social pressure and dietary behavior [46], further corroborating this study’s findings. To address these issues, a comprehensive, nurse-coordinated approach is required: (1) At the hospital level, a nurse-led educational system should be established to provide patients with personalized health guidance [47]. (2) At the community level: Group education programs should be implemented in communities to improve the treatment system [48]. (3) At the societal level: Community nurses help raise public awareness of diabetes, reducing the stigma associated with the condition, and addressing weight bias are areas that require attention in future healthcare efforts [45]. Insufficient perception of diabetes risk A participant exemplified this normalized risk-taking: “For example, let’s say I’m going out for drinks tonight. I inject a few extra units of insulin before I start drinking. Then I go out—everyone does it this way,” (Participant Q; female, 51–60 years, 11–20 years disease duration, urban residence, physical laborer) demonstrating how herd mentality erodes individual risk perception. This study’s coding revealed that the patients lacked awareness of potential warning signs. When patients observe other patients engaging in risky behaviors without suffering adverse consequences, they begin to imitate these actions, believing that doing the same will not result in any negative outcomes for themselves. This reflects optimism bias and survivor bias: [49, 50] observing only those “who got away with it,” while ignoring those who developed complications. Since diabetes is asymptomatic, disease perception relies on felt experience rather than medical warnings [51], patients find it difficult to develop a sense of the disease’s severity until they perceive a clear threat. This study also found that some patients did not focus on diabetes because it was masked by other prominent symptoms. To address this challenge, clinical nurses may improve patients’ risk perception by using visual risk presentations and decision-support tools, combined with discussions of the disease and its management tailored to each patient’s circumstances. [52, 53]. Lack of knowledge about diabetes One participant candidly admitted: “I basically know nothing about medical matters” (Participant E; male, 61–70 years, 6–10 years disease duration, rural residence, elementary school education), vividly illustrating the profound gap in health literacy. Beyond the generic concept of “eating less and exercising more,” patients lacked concrete self-management skills. More critically, this study revealed that patients faced a fundamental credibility crisis in knowledge acquisition: over half of participants could not effectively distinguish between reliable and unreliable online health information. As Participant H articulated: “I feel that information online is indeed helpful, but I remain skeptical about some of it because I suspect it might be false.” This dilemma stems not merely from insufficient health literacy but reflects a deeper structural problem—an information trust crisis in the digital age. Patients confronted with contradictory health advice from multiple sources became paralyzed by an inability to establish evaluative criteria. Participant O’s experience was particularly representative: “I watch a lot of health videos on my phone. Some say eat this, some say don’t eat that. The advice is all contradictory. I don’t know who to believe, so I just ignore most of it.“ [54] To address this structural barrier, nursing interventions should employ a two-tiered progressive strategy. First, at the knowledge-transmission level, interventions should move beyond unidirectional didactic teaching and adopt an interactive, patient-centered approach. Second, at the information evaluation level, nurses should equip patients with foundational information-assessment competencies—teaching patients to recognize authoritative sources (e.g., hospital official channels, evidence-based clinical guidelines) and distinguish them from non-authoritative sources. Recent research demonstrates that structured digital health literacy training significantly enhances patients’ capacity to critically appraise online health information [55]. Low self-efficacy Participants expressed significant anxiety regarding post-discharge self-management. As one stated: “Once I’m home, I certainly won’t have regular reminders to inject insulin or check my blood sugar as I do in the hospital,” (Participant L; male, 51–60 years, 1–5 years disease duration, urban residence, non-physical laborer), capturing the overwhelm of transitioning to autonomous home-based care. In our analysis, a significant portion of the codes reflected a deficiency in patients’ self-efficacy. Extensive research has demonstrated that higher self-efficacy is significantly associated with better self-management skills and more optimal glycemic control [23, 56, 57]. We identified two underlying psychological mechanisms—treatment fatigue and learned helplessness. Treatment fatigue arises from the physical and mental exhaustion of adhering to complex, long-term self-care routines, while learned helplessness reflects the disappointment that follows when sustained efforts fail to produce expected improvements. Both mechanisms have been documented in other fields of chronic disease [58, 59]. Therefore, nurses need to realize that enhancing self-efficacy requires not only knowledge-based approaches but also interventions focused on psychological factors and goal-setting. For example, helping patients set short-term, achievable goals allows them to see immediate results and gradually build self-management skills [60]. During follow-up, it is essential to monitor treatment fatigue and learned helplessness and provide necessary psychological support in a timely manner. High costs of management One participant epitomized the existential trade-off inherent in diabetes management: “It would be such a shame if I couldn’t eat what I want before I die. Everyone has to die sooner or later,” (Participant R; male, 71–80 years, > 20 years disease duration, urban residence, retired) articulating how strict dietary discipline conflicts with the search for life’s meaning. Beyond financial and practical burdens, diabetes management imposes an existential cost—the erosion of life’s meaning. The unremitting daily regimen of dietary restrictions, monitoring, and medication adherence consumes patients’ time, energy, and sense of purpose [61]. As the condition worsens, the financial burden on patients also increases [62]. However, an even more hidden cost is the sacrifice of the meaning of the remainder of one’s life in exchange for health management. This finding can be understood through the lens of existential psychology, which emphasizes the search for meaning in human existence. Frankl noted that existence is about finding meaning in suffering [63]. Research has found that some patients believe enjoying life and relishing food constitutes the meaning of existence. However, the onset of illness in patients is associated with their failure to maintain their previous lifestyle; if they fail to find a new sense of purpose in their lives during the self-management process, they will adopt a pessimistic attitude toward the remainder of their lives. To address this issue, implementing holistic care can be effective. Therefore, the healthcare system should promote the implementation of minimally disruptive medicine in diabetes management [64]. Meanwhile, community nurses should collaborate with the families of patients to help them rediscover the meaning and joy of life at home, thereby promoting sustainable self-management [65]. Finally, it is worth noting that internal and external rewards are not categorized separately but are instead scattered across other categories. When patients temporarily set aside their diabetes management in pursuit of culinary pleasure and social propriety, this reflects both a desire for short-term enjoyment and the need to maintain social relationships, while also highlighting the influence of self-efficacy and social support. Within China’s unique cultural context, these reward dimensions may be embedded within patients’ daily lives rather than manifesting as distinct, singular motivations. Theoretical contributions: refining PMT in a collectivist context The embedded nature of reward dimensions observed in this study prompts a critical reconsideration of Protection Motivation Theory [29]. Originally formulated in Western settings that tend to emphasize individualism, PMT posits that intrinsic and extrinsic rewards independently weaken willingness to engage in protective behavior. Our findings challenge this assumption: intrinsic and extrinsic rewards did not emerge as separate categories, but remained implicit and embedded within the social fabric of participants’ experiences, most clearly under “lack of social understanding and care.” In a collectivist culture, maintaining face (Mianzi), preserving social harmony at family meals, and honoring social obligations toward friends can function as potent external motivators. For these Chinese patients, avoiding social embarrassment or exclusion frequently outweighed appraisals of personal health threats. By showing that collectivist cultural norms can erode both risk perception and self-efficacy, our findings refine PMT. We therefore propose that when applying PMT in collectivist or other non-Western settings, “rewards” should be interpreted through a broader psychosocial lens that acknowledges the significance of social conformity. Limitations Limitations of this study warrant acknowledgment. First, all participants were hospitalized during interviews, which may select for patients with more severe complications or complex care needs; consequently, identified barriers (particularly high perceived costs and low self-efficacy) may be more pronounced than in community-based samples. Second, participants were recruited from a single tertiary hospital in Northeast China, limiting generalizability to other regions or healthcare systems. Third, social support was assessed only via marital status, which inadequately captures operative support dimensions (e.g., living arrangements, caregiver presence, actual support provision). Future research should incorporate more comprehensive social support measurement. Fourth, interview duration (approximately 20 min; range: 15–35 min) may have limited depth of exploration for complex categories such as existential costs and learned helplessness. Fifth, as a qualitative study using retrospective recall, results are subject to recall bias and social desirability bias, despite efforts to establish rapport and ensure anonymity. Conclusion Diabetes self-management failure in Chinese hospitalized patients results from interacting barriers at cognitive, psychosocial, and environmental levels. Using Protection Motivation Theory, we identified six primary categories: environmental stress, insufficient social support, low risk perception, knowledge gaps, low self-efficacy, and high management costs. Key Recommendations for Clinical Practice: Nurses should conduct a systematic pre-discharge assessment addressing both threat and coping appraisals—using visual risk tools to enhance disease perception, exploring confidence in specific self-management strategies, and identifying individual barriers and strengths. A culturally tailored, nurse-led transitional care model that bridges hospital, community, and home settings is essential. By honoring patients’ personal values alongside health priorities and providing ongoing psychological support, this approach can strengthen self-efficacy and sustain long-term adherence to self-management behaviors. The recommended pre-discharge assessment checklist is presented in Table 5. Table 5Nurse-led pre-discharge assessment checklist based on PMT constructsAssessment DomainPMT ConstructKey QuestionsTime AllocationThreat Appraisal (Step 1: Priority)Perceived Severity & Susceptibility• Can patient name at least 2 diabetes complications?• Does patient verbalize personal susceptibility to these complications?• Rating: Patient’s perceived risk level (1–10 scale)5 minCoping Appraisal (Step 2: Priority)Self-Efficacy & Response Efficacy• How confident is patient in specific post-discharge behaviors (e.g., insulin injection, dietary adherence, blood glucose monitoring)?• Does patient believe these actions will be effective?• Confidence rating for each behavior (1–10 scale)5 minBarrier Identification (Step 3: Priority)Perceived Costs & Environmental Factors• What specific obstacles exist at home (family dynamics, work demands, financial constraints, social pressures)?• Which barriers are most likely to compromise adherence?• Identify 1–2 modifiable barriers for targeted intervention5 minWhen time is severely limited (< 15 min), prioritize assessments in this sequence:1. First: Threat Appraisal (5 min) – Establish baseline disease risk perception; use brief visual aids (e.g., complication severity charts)2. Second: Coping Appraisal (5 min) – Assess confidence in specific post-discharge behaviors; identify skill gaps3. Third: Barrier Exploration (5 min) – Identify 1–2 modifiable environmental or psychosocial obstacles; develop targeted mitigation strategies Nurse-led pre-discharge assessment checklist based on PMT constructs • Can patient name at least 2 diabetes complications? • Does patient verbalize personal susceptibility to these complications? • Rating: Patient’s perceived risk level (1–10 scale) • How confident is patient in specific post-discharge behaviors (e.g., insulin injection, dietary adherence, blood glucose monitoring)? • Does patient believe these actions will be effective? • Confidence rating for each behavior (1–10 scale) • What specific obstacles exist at home (family dynamics, work demands, financial constraints, social pressures)? • Which barriers are most likely to compromise adherence? • Identify 1–2 modifiable barriers for targeted intervention When time is severely limited (< 15 min), prioritize assessments in this sequence: 1. First: Threat Appraisal (5 min) – Establish baseline disease risk perception; use brief visual aids (e.g., complication severity charts) 2. Second: Coping Appraisal (5 min) – Assess confidence in specific post-discharge behaviors; identify skill gaps 3. Third: Barrier Exploration (5 min) – Identify 1–2 modifiable environmental or psychosocial obstacles; develop targeted mitigation strategies Recommended Pre-Discharge PMT-Based Assessment Framework (15-minute format for time-constrained clinical settings): This framework operationalizes the study’s theoretical findings into actionable clinical assessment, enabling nurses to identify and address the most critical barriers within typical hospital discharge timelines. Supplementary Information Below is the link to the electronic supplementary material. Supplementary Material 1 Supplementary Material 2 Supplementary Material 3 Acknowledgements We would like to express our gratitude to the patients who participated in this research, as well as the researchers. Abbreviations Type 2 diabetes mellitus Protection Motivation Theory American Diabetes Association Health Belief Model Behavior Change Wheel Consolidated criteria for reporting qualitative research Author contributions All authors who participated in this research made significant contributions: ZYX proposed the topic of the paper, participated in the formulation of the interview outline, conducted pre-interviews and formal interviews, processed the data, discussed the research results, and wrote and revised the first draft of the paper. TL was responsible for designing the interview outline, training for semi-structured interviews, seeking financial support, discussing the research results, and reviewing the first draft of the paper. YL was responsible for designing the interview outline, final review of the research results, and reviewing the first draft of the paper. RQC and SYY were responsible for discussing the research results and reviewing the first draft of the paper. Meanwhile, RQC also participated in data processing. All authors read and approved the final manuscript. Funding Liaoning Provincial Science and Technology Program (Grant No. 2025-MSLH-240). Data availability The interview transcripts generated and analyzed during the current study are not publicly available due to concerns about participant confidentiality but are available from the corresponding authors on reasonable request and subject to institutional and ethical approval. Declarations Ethics approval and consent to participate This study was approved by the Ethics Committee of Jinzhou Medical University (Approval No. JZMULL2026004) and conducted in accordance with the Declaration of Helsinki. Written informed consent was obtained from all participants prior to the interviews. Participants were informed that they could withdraw from the study at any time without affecting their medical care. Consent for publication Not applicable. Competing interests The authors declare no competing interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Contributor Information Tao Liu, Email: 534575061@qq.com. Ying Liu, Email: liuy2@jzmu.edu.cn. References Associated Data Supplementary Materials Supplementary Material 1 Supplementary Material 2 Supplementary Material 3 Data Availability Statement The interview transcripts generated and analyzed during the current study are not publicly available due to concerns about participant confidentiality but are available from the corresponding authors on reasonable request and subject to institutional and ethical approval.