Introduction
Type 2 diabetes mellitus (T2DM) is one of the most common chronic metabolic diseases in China. In 2023, the age-standardised prevalence of diabetes in China reached 13.67%, with T2DM accounting for more than 90% of all diabetes cases. Its prolonged disease course and substantial burden of complications have become a major public health concern.1 Evidence indicates that T2DM is particularly prevalent among middle-aged and older adults.2 This population commonly experiences progressive declines in physical, psychological, and cognitive functioning and often has multiple comorbidities, making them more susceptible to poor glycaemic control and marked glycaemic variability and, consequently, to a higher rate of unplanned readmission.3 Unplanned readmission refers to an unexpected rehospitalisation due to deterioration in a patient’s condition within a period following the initial discharge.4 Studies have shown that the rate of unplanned readmission among patients with diabetes ranges from 14% to 20% after discharge,5 and inadequate health behaviour management further increases the risk of readmission.6 A recent prospective study in China reported an all-cause readmission rate of 18.6% within 90 days after discharge among patients with diabetes.7 In the present study, repeated hospitalisations were operationally defined as at least two unplanned hospital admissions within the preceding three months, including the index admission.
Effective health behaviour management helps patients with T2DM achieve glycaemic control and reduce the risk of complications. Changes in health behaviour, however, depend on sound health behaviour decision-making, which is shaped by multiple factors, including individual cognition, real-life circumstances, and previous experiences. Whether patients can make informed and appropriate decisions about their health behaviours directly affects the effectiveness of disease management.8,9 For patients who experience repeated hospitalisations, recurrent admissions often indicate greater complexity in disease management, and the associated health behaviour decision-making is therefore more dynamic and context dependent.
Although the importance of health behaviour management in T2DM has been widely recognised, existing research has been predominantly quantitative, focusing mainly on current health behaviours and their influencing factors, with limited in-depth exploration of the health behaviour decision-making process among patients with T2DM who experience repeated hospitalisations.10,11 Notably, although previous qualitative studies have explored self-management behaviours and care-seeking decisions among patients with T2DM,12,13 the stage-specific health behaviour decision-making of patients experiencing repeated hospitalisations, particularly across the transition from hospitalisation to home management, remains insufficiently understood. If clinical understanding of this continuous process is limited, decision support strategies may become fragmented, thereby weakening the targeting and effectiveness of disease management interventions.
Patient journey mapping, as a patient-centred methodological tool, offers a useful approach to addressing the limitations of conventional methods in understanding patient experiences and healthcare system processes.14 Originating in service design, it uses visual narrative strategies to longitudinally integrate patients’ behavioural trajectories, emotional responses and evolving needs across the care continuum, thereby revealing the dynamic interactions between patient experiences and care processes more systematically.15 Unlike research paradigms focused on isolated variables or single time points, patient journey mapping helps identify critical turning points in patient experiences, detect discontinuities in care processes and translate patient narratives into actionable intervention targets.16 Existing studies have applied patient journey mapping to patient experience research, patient-centred care for chronic conditions and service process optimisation.17–20 By mapping patient journeys, healthcare systems can more clearly identify patient-centred care patterns, structural barriers and gaps in system performance, thereby translating these insights into implementable interventions.
Accordingly, this study adopted a longitudinal qualitative design based on patient journey mapping. Repeated semi-structured interviews were conducted with patients with T2DM who experienced repeated hospitalisations to explore the evolution of their health behaviour decision-making across the pre-admission, hospitalisation, and home-management stages. The study sought to identify stage-specific behavioural patterns, support needs, emotional experiences, and pain points and to develop a journey map of health behaviour decision-making, thereby providing a basis for designing stage-specific decision-support strategies and continuity-of-care interventions for this population.
Methods
Study Design
The patient journey map was constructed through a three-stage approach: (1) an initial journey framework was outlined by drawing on relevant literature and clinical diagnosis and treatment processes; (2) key touchpoints and experience nodes were further clarified through clinical observation; and (3) details of patients’ decision-making experiences were enriched and potential intervention opportunities were identified through repeated semi-structured interviews.21 The final map was validated through collaborative evaluation with key stakeholders, including physicians, nurses, patients. This study adopted a descriptive qualitative approach to explore the complex lived experiences of health behaviour decision-making among patients with T2DM experiencing repeated hospitalisations. It aimed to provide actionable insights for clinical practice rather than to generate a new theoretical model.
This approach is particularly suited to research areas where direct description of phenomena is required.22 Through repeated semi-structured interviews and content analysis, this study design enabled patients to provide relatively holistic accounts of their health behaviour decision-making experiences, thereby revealing the detailed processes through which decisions were formed, implementation was hindered and support needs changed across different contexts.
Desktop Review Research
The research team used a desktop review approach to systematically integrate literature evidence, clinical care materials and disease management information concerning health behaviour decision-making among patients with T2DM and a history of repeated hospitalisations.21,23 Through repeated comparison and team discussions, we developed a timeline of patients’ care journeys and divided it into 3 key stages: (1) pre-admission (2) in-hospital treatment and (3) home management. Key decision points and management transitions related to health behaviour decision-making were then systematically incorporated into the initial visual framework to guide the development of the subsequent interview guide and patient journey map. The seven care-process substages presented in Figure 1 were identified through the literature and clinical care-process review and refined through preliminary clinical observations.
Figure 1Journey Map of Health Behaviour Decision-making Among Patients with T2DM Experiencing Repeated Hospitalisations.A journey map arranged as a timeline with three stage headers: Pre-admission, Hospitalisation and Home Management. Stage row with seven substages. Under Pre-admission: Diagnostic confirmation and Preparation for Hospitalisation.
Under Hospitalisation: Acute Phase, Stabilisation Phase and Discharge Preparation. Under Home Management: Home management and Long-term adaptation. Behaviour row.
Pre-admission section header reads, preference-driven care-seeking decisions, with three items: Hesitant initiation of care-seeking; Familiarity-driven choice of healthcare institution; Trade-off decision-making under practical. Hospitalisation section header reads, Structured Treatment Adherence, with three items: Following Instructions Under Healthcare Professional-led care; Standardised Behavioural Regulation; Reassurance-based Dependence in Decision-making. Home Management section header reads, fragile self-management, with three items: Independent maintenance of self-management routines; Regular follow-up visits; Experience-based self-adjustment.
Needs row. Pre-admission section header reads, seeking trusted decision anchors, with three items: Seeking Authoritative Information; Assessment of Treatment Suitability; Reliance on Trusted Reference Points. Hospitalisation section header reads, Seeking Structured Empowerment and Transition Preparation, with three items: Understanding the Treatment Plan; Guidance on Self-management Skills; Preparation for the Transition From Hospital to Home.
Home Management section header reads, Seeking Continuous and Accessible Out-of-hospital Support, with three items: Ongoing Follow-up Support; Convenient Access to Consultation Services; Timely Professional Responses. A row of overarching statements aligned to the timeline reads: Self-blame After Initially Neglecting the Illness; Fragile Stability; Acceptance and coexistence with the Disease Under Long-term Burden. Emotions row shows a curved line with seven marked points and callout labels, from left to right: Underestimation of Illness Severity; Regret and Self-blame; Frustration Associated with Readmission; Temporary Sense of Security; Persistent Worry; Fatigue and depletion; Acceptance and adjustment; Coexistence and inner calm.
Pain points row. Pre-admission section header reads, Decision-Making Amid Multiple Information Dilemmas, with three items: Inadequate Health Literacy; Information Overload; Low Willingness to Seek Care. Hospitalisation section header reads, Insufficient Motivation for Sustained Self-management, with three items: Insufficient Reflection on the Causes of Readmission; Concerns About Sustained Management Investment; Insufficient Recognition of Disease Severity.
Home Management section header reads, A Weakened Out-of-hospital Support Ecosystem, with three items: Unstable Family Support; Burden of Long-term Management; Decision-making Pressure in Social Contexts.A journey map of stages, behaviours, needs, emotions and pain points for repeated hospitalisations.Note: The emotional curve schematically represents participants’ reported emotional experiences across care stages and is not based on quantitative emotion measurements.
Participatory Observation Method
Structured observations were conducted by 3 registered nurses with experience in diabetes care with 3 patients with T2DM experiencing repeated hospitalisations who had complete experiences across the care journey. The nurses recorded, in detail, hospital-based interactions related to health behaviour decision-making. The observational field notes documented patient–healthcare professional interactions related to treatment and self-management decisions, including patients’ questions, expressed concerns, and responses to professional guidance.
Based on a systematic analysis of the observational data, the research team constructed and refined the initial journey framework and specified 4 core dimensions within the health behaviour decision-making journey (decision-making behaviours, support needs, emotional responses and pain points). The three patients involved in the preliminary observations were not included in the final longitudinal sample.
Participants
Participants were recruited between October and December 2025 from the Department of Endocrinology at a tertiary Grade A hospital in Shenzhen, China, using maximum-variation purposive sampling. The inclusion criteria were as follows: (1) aged ≥18 years; (2) met the diagnostic criteria for T2DM,24 and had experienced at least two unplanned hospital admissions within the preceding three months, including the index admission, due to poor glycaemic control or diabetes-related conditions (acute diabetes-related complications, progression of chronic diabetes-related complications, or diabetes-related infections), and (3) were able to understand and complete the interview. The exclusion criteria were as follows: (1) admission for scheduled follow-up examinations, routine health checks, or other planned assessments; (2) pregnancy; and (3) the presence of another serious illness.
The dates, frequency, unplanned status, and principal reasons for hospital admissions within the preceding three months, including the index admission, were verified through the hospital’s electronic medical records. The reasons for admission were categorised with reference to the Agency for Healthcare Research and Quality (AHRQ) Prevention Quality Indicators (PQI), version 2025,25 and the Guideline for the Management of Diabetes Mellitus in the Elderly in China (2024 Edition).26 The sample size was determined according to the principle of thematic saturation. Data collection and analysis were conducted concurrently.
After each interview, the research team summarised the interview and conducted preliminary coding. After analysis of data from the 14th participant, no new core themes emerged. To rigorously confirm saturation and reduce the potential impact of attrition in the repeated-interview design on the completeness of findings, 2 additional participants were recruited.
The additional data were accommodated within the existing thematic framework and generated no new thematic categories, indicating that thematic saturation had been reached. The final sample comprised 16 participants, and 47 interviews were completed. One participant missed the home management stage interview 3 months after discharge.
Research Instruments
The General Social Demographic Questionnaire
This study used a general social demographic questionnaire developed through a literature review and team discussions to collect sociodemographic and clinical data, including age, sex, educational level, place of residence, employment status, duration of T2DM, complications, living arrangements and self-reported health status. Self-reported health status was collected through a brief descriptive background item using the following single-item self-rated health prompt: How would you rate your overall health at present? This item was used to describe sample characteristics, support maximum variation purposive sampling and contextualised interpretation of participants’ narratives.
Within this qualitative study design, this item was not treated as an outcome measure and was not subjected to psychometric testing.
The Longitudinal Semi-Structured Interview Outline
The interview guide was developed based on the study objectives, literature review and findings from group discussions. After pilot interviews with 3 patients were completed, the guide was revised and refined. Two experts in qualitative research and diabetes-related fields were invited to review the interview guide, and its content was further refined according to their feedback.
The semi-structured interview guide was designed to systematically trace the key stages from symptom onset to long-term adaptation. Each question focused on core dimensions that were critical for mapping patients’ experiences, These pilot interview participants were not included in the final longitudinal sample. Table 1.Table 1Semi-Structured Interview Guide and Target ConstructsInterview QuestionsTarget Constructs1.
How do you usually assess changes in your condition?Illness perceptionRisk recognitionIllness attribution2. At what point do you decide to seek hospital admission and treatment? Do you make this decision yourself?
What factors influence your decision?Triggers for the hospitalisation decisionDecision-making autonomyFactors influencing decision-making3. What healthcare decisions have you been involved in during hospitalisation? How did you feel when making these decisions?Participation in decision-making during hospitalisationPatient–healthcare professional interactionsEmotional responses4.
What difficulties have you encountered when making decisions? How did you cope with them?Decision-making difficultiesPractical barriersCoping strategies5. What support would you like to receive from healthcare professionals, family members, or others?Support needsUnmet needsStakeholder support6.
Looking back on your repeated hospitalisations, how has your health behaviour decision-making changed?Changes in decision-makingInfluence of hospitalisation experiencesChanges in self-management
Research Team
The study was conducted by an interdisciplinary team comprising three lead researchers (H M, M Z, and WX H), four research assistants (WL Y, BX Y, T L, and YX L), and two supervisors (MJj L and WW C). All team members had received formal training in qualitative research methods, and the team included clinical nurses experienced in chronic disease management and diabetes care. The three lead researchers were master’s students in nursing.
They were responsible for data collection: one led the interviews, one documented field observations and non-verbal information, and the third organised the collected data. The interviewer had not been involved in providing clinical care to the participants. The research assistants were responsible for transcript verification, initial coding, and theme development.
The supervisors conducted regular validity checks through peer-debriefing sessions to resolve any ambiguities in the interpretation of the data during the analytical process. During peer-debriefing sessions, the research team critically examined how their clinical experience and prior assumptions might influence the interpretation of participants’ accounts.
Data Collection
Repeated semi-structured interviews were conducted at three time points: on the second day of hospitalisation (T1), one day before discharge (T2), and three months after discharge (T3). The first interview retrospectively explored participants’ pre-admission experiences.
Interviews were conducted in quiet and comfortable settings, with the specific location determined according to each participant’s preference. Before the interview, the researchers explained the study purpose, expected duration, and confidentiality measures and obtained permission to audio-record the conversation. Interviews commenced and were audio-recorded after participants had provided written informed consent.
Throughout each interview, the researcher listened attentively and documented participants’ body movements, facial expressions, emotional fluctuations, and subtle changes in tone of voice. Key responses were verified using qualitative interviewing techniques, including probing, paraphrasing, and clarification, to ensure data accuracy and a comprehensive representation of participants’ perspectives. The sequence of interview questions was adjusted flexibly according to the flow of the conversation, and data collection continued until data saturation was reached.
Each interview lasted 30 to 60 minutes. To protect participants’ privacy, all identifying information was replaced with coded identifiers.
Data Analysis
Interview data were analysed using a framework-guided directed qualitative content analysis.27 The four dimensions established through preliminary patient observations served as the initial analytical framework. Interviews were conducted in Mandarin Chinese, and audio recordings were transcribed and organised in Chinese within 24 hours of each interview. Coding was conducted in Chinese; selected quotations were translated into English and checked against the original Chinese transcripts.
Researchers reviewed each transcript repeatedly and extracted relevant statements to generate initial codes. Text segments that did not fit the initial framework were assigned new codes, which were subsequently examined for their relationship to the four established dimensions. The resulting codes were then interpreted and grouped into subthemes to characterise stage-specific experiences within the four dimensions.
For longitudinal analysis, interviews were organised by participant and time point to compare changes and continuities within participants across T1–T3, followed by cross-case synthesis of shared and divergent patterns. The available T1 and T2 interviews from the participant who missed T3 were retained in the analysis.
Using WPS Office (version 12.1.0.28505), the journey map was organised and visualised as a sequential flow diagram. Patient journey mapping was not used as an analytical method, but as an integrative framework and visualisation tool for organising and presenting the findings. The map was developed after data analysis by integrating the identified key themes and categories and arranging them chronologically across the continuum of care.
Two participants who had completed all three interviews and were able to clearly articulate their experiences reviewed the preliminary thematic classifications and draft journey map to assess whether the findings reflected their experiences. Three diabetes specialist nurses, one professor with clinical expertise in endocrinology and one professor specialising in qualitative research reviewed the coding process and draft journey map. Feedback prompted minor refinements to the content and presentation of the journey map, primarily involving the wording of selected themes, to better reflect participants’ experiences.
Rigour
The Con-solidated Criteria for Reporting Qualitative Research (COREQ) guidelines were used to evaluate the qualitative research reports in this study.28 To enhance credibility, triangulation was incorporated into the data analysis through investigator analysis, participant validation, and expert review. Detailed records of the interview guide, data transcription, and analytical procedures were maintained throughout the study to ensure auditability and methodological dependability. As the analysis progressed, additional interviews yielded only limited substantive insights relevant to the dimensions of the journey map, supporting the analytical sufficiency of the data in relation to the study aims.
Transferability was enhanced through a rich description of health behaviour decision-making among patients with T2DM who experienced repeated hospitalisations. Interviewing techniques, including attentive listening, in-depth probing, paraphrasing, and clarification, were applied systematically to reduce potential social desirability bias in participants’ responses.
Ethical Considerations
This study was performed in accordance with the principles of the Declaration of Helsinki. All participants provided written informed consent after receiving information about the purpose and significance of the study. Participant privacy and data confidentiality were strictly protected throughout the study.
All audio recordings and interview transcripts were used solely for the purposes of this research and were not disclosed or reused without explicit consent. Ethical approval was granted by the Institutional Review Board of Shenzhen Hospital of Guangzhou University of Chinese Medicine (Futian) (Approval No. GZYLL(KY)-2025-108).
Results
Overall Sociodemographic Characteristics of the Participants
This longitudinal qualitative study included 16 patients with T2DM who had experienced repeated hospitalisations. The sample comprised eight men and eight women. Participants ranged in age from 46 to 72 years (59.38 ± 8.07), and the duration of diabetes ranged from 2 to 25 years (10.47 ± 7.65).
Regarding educational attainment, 62.5% of participants had received higher education. In addition, 81.3% lived in urban areas, 56.3% were retired, 50.0% lived with a spouse, and more than 90% had fewer than three complications. Regarding hospitalisation history, 75.0% of participants had experienced two unplanned hospital admissions and 25.0% had experienced three within the preceding three months.
Among the 36 recorded admissions, poor glycaemic control (55.6%) and progression of chronic diabetes-related complications (27.8%) were the most common reasons for admission. With respect to self-rated health, 43.8% of participants rated their overall health as fair (Table 2).Table 2The General Sociodemographic Characteristics of the Participants (n=16)VariableCategoryFrequency (n)Percentage (%)Age40~50212.551~60743.8>60743.8SexMale850.0Female850.0Educational levelPrimary education212.5Secondary education425.0Higher education1062.5Place of residenceRural318.8Urban1381.3Employment statusUnemployed16.3Employed637.5Retired956.3Duration of diabetes (years)<10850.010~15425.0>15425.0Living arrangementLiving alone212.5Living with children637.5Living with a spouse850.0Number of complications0531.31531.32531.3≥316.3Number of unplanned hospital admissions in the preceding 3 months21275.03425.0Principal reasons for unplanned hospital admissionsPoor glycaemic control2055.6Acute diabetes-related complications38.3Progression of chronic diabetes-related complications1027.8Diabetes-related infections38.3Self-reported health statusPoor531.3Fair743.8Good425.0Notes: Self-reported health status was assessed using a brief descriptive background question with the following single-item prompt: “How would you rate your current overall health?” This item was used to describe the sample characteristics and to support maximum-variation sampling and the contextual interpretation of participants’ narratives.
Development of the Journey Map for Patients with T2DM Experiencing Repeated Hospitalisations
Figure 1 presents a patient journey map that integrates the evolving experiences of patients with T2DM who experienced repeated hospitalisations across three stages of health behaviour decision-making: pre-admission, hospitalisation, and home management. The journey map provides a longitudinal, patient-centred perspective on how patients adapted over time and highlights specific opportunities to improve service delivery and quality of care at each stage.
The longitudinal interviews yielded 35 initial codes, which were grouped into 12 subthemes and used to characterise stage-specific trajectories within the four previously established dimensions: decision-making behaviour, support needs, emotions, and pain points. Together, these stage-specific findings captured changes in patients’ health behaviours, needs, emotional responses, and pain points, as well as key sources of friction that undermined patient engagement and decision quality. The integrated findings were incorporated into the journey map to identify specific targets for nurse-led interventions and continuity support at each stage.
Illustrative patient quotations are presented in Supplementary File 1.
Longitudinal Trajectories Aligned with the Journey Map
Behaviour
Behavioural trajectory: Patients’ behavioural trajectory evolved from preference-driven care-seeking decisions before admission, to structured adherence under healthcare professional guidance during hospitalisation, and ultimately to fragile self-management during home management, characterised by autonomous maintenance within daily contexts but vulnerability to disruption.
Pre-admission: Preference-driven care-seeking decisions. Patients often delayed initiating healthcare-seeking decisions after repeated observation and hesitation. Even when they decided to seek medical care, their choices were frequently shaped by familiarity, previous healthcare experiences, and practical circumstances, rather than being based on consistent proactive recognition of health risks.
P10: At that time, I was unsure whether I should come [to the hospital]. Also, I had to take care of my child at home, so I thought I could hold on for a while. Later, my blood glucose became higher and I felt uncomfortable, so I came to the hospital immediately.
(Supplementary File 1, No. 1, Quotation 1)P12: Mainly because I am familiar with this place, and the doctor knows my condition. I feel more reassured coming here.
(Supplementary File 1, No. 2, Quotation 1)
Hospitalisation: Structured treatment adherence. During hospitalisation, patients primarily followed standardised care arrangements and medical instructions. Their behaviours were characterised by adherence to predetermined treatment plans led by healthcare professionals, with some aspects of decision-making delegated to professional providers.
P6: Someone monitors me and someone manages me. When I stay here, I listen to the doctors and nurses. I don’t have to think too much on my own, and I feel more at ease.
(Supplementary File 1, No. 4, Quotation 1)P15: After being admitted, the schedule, meals, and medication use are much more regular. Nurses also remind me, so it is not easy to become lazy even if I want to.
(Supplementary File 1, No. 5, Quotation 1)
Home management: Fragile self-management. After discharge, patients began to manage their condition more independently. In the absence of continuous professional supervision, some patients attempted to maintain established management routines, while others increasingly relied on personal judgement and previous experience to adjust their management.
P5: The doctor told me to return every week during the first month after discharge and, once things became stable, to come once a month. I followed that schedule strictly. With this illness, you need a doctor to keep an eye on you.P7: If my blood glucose gets too high, I take a little extra medication myself.
I’ve done it before, and nothing happened.
Decision-Support Needs
Decision-support trajectory: Patients’ decision-support needs evolved from seeking trusted decision anchors before admission, to seeking structured empowerment and transition preparation during hospitalisation, and ultimately to seeking continuous and accessible out-of-hospital support during home management.
Pre-admission: Seeking trusted decision anchors. When confronted with abundant information and multiple treatment options, patients urgently sought professional guidance that could help them judge the reliability of available choices and determine whether specific treatments were appropriate for their own conditions. Their core need was not simply to obtain more information, but rather to identify a trusted reference point upon which they could base their decisions.
P8: There is so much online information, like Chinese medicine or Western medicine that claims to cure diabetes. I bought many of them and spent several thousand yuan in total. (Supplementary File 1, No.
28, Quotation 1)P9: I was always thinking whether there was a faster way. I tried many medicines recommended online or by friends, but they did not work. After going around in circles, I still had to come to the hospital.
(Supplementary File 1, No. 12, Quotation 2)
Hospitalisation: Seeking structured empowerment and transition preparation. During hospitalisation, patients hoped to receive more specific management guidance while undergoing standardised treatment, enabling them to better understand their current treatment plans and prepare for ongoing management after discharge. Their needs extended beyond acquiring knowledge during hospital care and included practical, actionable support to facilitate the transition from hospital-based care to home management.
P3: I need the doctor to explain it to me. How should I take this medication? How should I give myself this injection?
(Supplementary File 1, No. 14, Quotation 1)P11: My complication is quite serious now. The doctor said I needed to stop metformin before surgery and switch to insulin.
After discharge, when should I switch back to metformin? (Supplementary File 1, No. 13, Quotation 1)
Home management: Seeking continuous and accessible out-of-hospital support. During home management, patients were more concerned about whether they could continue to receive professional feedback and timely responses after leaving the hospital environment. Their needs were primarily reflected in the expectation that out-of-hospital support could be maintained over time and that professional assistance could be easily accessed when problems actually occurred.
P1: Previously, the community healthcare doctor would call me and say, ‘Auntie, it is time for you to come and have your blood glucose checked.’. He arranged a check once a month and also asked whether I had run out of medication. He was very responsible, and I found this very helpful.
(Supplementary File 1, No. 16, Quotation 1)P16: It does not necessarily have to be face-to-face. But when I really have a problem, I hope I can find someone to ask more easily.
(Supplementary File 1, No. 17, Quotation 1)
Emotions
Emotional trajectory: Patients’ emotional trajectory evolved from self-blame after initially neglecting the illness before admission, to fragile stability within the hospital care environment, and ultimately to acceptance and coexistence with the disease under long-term burden.
Pre-admission: Initial underestimation of the illness followed by self-blame. Some patients underestimated the severity of their condition and frequently attributed their symptoms to the natural ageing process, failing to associate them with disease progression. As their condition deteriorated, these patients often experienced feelings of regret and self-blame.
P8: I didn’t think it was a big problem. Later I thought, at our age, that’s just normal. (laughs) (Supplementary File 1, No.
19, Quotation 1)P5: It all comes down to my appetite. If I had controlled it earlier, I probably wouldn’t have ended up in the hospital again. (Supplementary File 1, No.
20, Quotation 1)
Hospitalisation: Fragile stability. Repeated hospitalisation often left patients feeling discouraged by the limited effectiveness of their day-to-day disease management. At the same time, the structured treatment environment and continuous monitoring by healthcare professionals provided a temporary sense of security.
However, this sense of stability did not alleviate their persistent concerns about glycaemic control after discharge, the progression of complications, or losing control of their condition again. P7: If my condition stayed stable, who would want to keep coming back to the hospital? Things get better after each adjustment, but after being home for a while, they become unstable again.
After this happens several times, you just start to feel discouraged. (Supplementary File 1, No. 21, Quotation 1)P6: The doctors make their rounds every day.
If I feel even a little uncomfortable, I can tell them straight away. I don’t have to worry about anything—just wait for the treatment. (Supplementary File 1, No.
22, Quotation 1)P13: I’ve had this numbness in my feet for quite some time. The doctor told me that keeping my blood glucose under control could slow its progression, but I still worry that it will keep getting worse. (Supplementary File 1, No.
23, Quotation 1)
Home management: Acceptance and coexistence with the disease under long-term burden. Patients gradually came to accept that diabetes requires lifelong management and must be lived with. However, this acceptance was neither immediate nor effortless.
Rather, it developed alongside the accumulated fatigue and emotional exhaustion associated with sustained self-management. Against this backdrop, some patients began to adopt a more settled attitude towards living with the disease over the long term. P4: I’ve already been very careful.
I eat less rice, go for a walk every day, and never miss my medication. But my blood glucose seems to work against me. Recently, the numbness in my feet has started again, and I know things may be getting worse.
It’s not just sadness—it’s exhaustion, emotional exhaustion. (sighs) (Supplementary File 1, No. 24, Quotation 1)P1: Once you accept it, you realise this disease is going to stay with your body.
You just have to learn to live with it peacefully. (Supplementary File 1, No. 26, Quotation 1)
Pain Points
Pain-point trajectory: Patients’ pain points evolved from Impeded decision-making amid multiple information dilemmas before admission, to insufficient motivation for sustained self-management during hospitalisation, and ultimately to a weakened out-of-hospital support ecosystem during home management.
Pre-admission: Impeded decision-making amid multiple information dilemmas. Patients often struggled to judge the credibility of health information and the suitability of treatment options because of limited health literacy, while information overload further undermined their ability to make consistent decisions. At the same time, some patients showed limited willingness to seek medical care and were more likely to delay treatment or deviate from more appropriate management pathways in the absence of clear professional guidance.
P2: Last year, I came across a lecture on WeChat claiming that diabetes could be reversed in three days with a traditional herbal remedy. It seemed to work while I was taking it, but as soon as I stopped, my blood glucose rebounded to an even higher level. Later, I found out that it might actually have contained Western medicine.
(Supplementary File 1, No. 27, Quotation 1)P15: A few months ago, my fasting blood glucose reached 9 mmol/L. I thought eating less rice and more vegetables would be enough, but I only kept it up for three days.
Then I got busy travelling for work and started ordering takeaways again. I didn’t know how I should adjust things, so I simply gave up. (Supplementary File 1, No.
28, Quotation 2)
Hospitalisation: Insufficient motivation for sustained self-management. During hospitalisation, care was largely centred on managing current symptoms and optimising glycaemic control, whereas the underlying management gaps contributing to repeated hospitalisations and the direction of long-term disease management were rarely addressed. Patients’ pain points were reflected in their limited understanding of why they had been readmitted, insufficient awareness of the long-term risks of diabetes, and continual weighing of the value of sustained self-management against their own capacity to maintain it.
P7: After I was admitted, the doctors mainly focused on my swollen and painful foot and on adjusting my blood glucose, and they told me what to do after I went home. But no one explained why my condition had worsened to the point that I needed to be admitted again. (Supplementary File 1, No.
30, Quotation 1)P1: My sons all have to work. It’s only diabetes, so I don’t want them to worry too much about me. (Supplementary File 1, No.
32, Quotation 1)P8: If the patches and moxibustion I paid for don’t make any obvious difference, I really feel like I’m wasting my money. (Supplementary File 1, No. 31, Quotation 1)
Home management: A weakened out-of-hospital support ecosystem. Patients’ pain points became more apparent during long-term home management. Inconsistent family support, the cumulative burden of ongoing self-management, and practical pressures arising from work and everyday life all influenced their decision-making, making it more difficult to sustain self-management behaviours over time.
P5: No one reminds me. My daughter doesn’t pay much attention either. She’s young, and she doesn’t really worry about these things.
(Supplementary File 1, No. 33, Quotation 1)P16: After discharge, the nurses would call to follow up, but they usually just asked how I was recovering and reminded me to take my medication and watch for symptoms. It was difficult to have a more in-depth discussion.
But if the follow-up was changed to face-to-face visits, that wouldn’t really be practical either. (Supplementary File 1, No. 34, Quotation 2)P12: I install air conditioners on a contract basis.
There’s no basic salary—we only get paid when we work. My boss keeps my schedule packed every day, so it’s impossible to take my medication or meals on time, and I can’t even get proper rest. (Supplementary File 1, No.
35, Quotation 1)
Discussion
This study developed a journey map of health behaviour decision-making among patients with T2DM who were repeatedly hospitalized, providing a deeper understanding of their dynamic decision-making process from the perspectives of both patient experience and care optimisation. The journey map illustrated stage-specific patterns in behavioural decision-making, decision-support needs, emotional responses, and key pain points across different stages of care. This longitudinal, stage-based visualisation suggests that health behaviour decision-making varied across care stages rather than following a uniform progression towards improvement.
Rather, it is continuously shaped by the interplay of risk recognition, external support, real-life circumstances, and the long-term burden of disease management. Therefore, improving the quality of disease management requires not only the identification of the predominant decision-making barriers and emotional challenges at each stage, but also the provision of targeted, stage-specific support at key decision points. From the perspective of care providers, the journey map captures the complex dynamics and multidimensional interactions inherent in patients’ health behaviour decision-making.
It may help healthcare professionals identify critical gaps in decision support across hospital and home settings, recognise weaknesses in service provision, and inform the development of patient-centred, comprehensively integrated service-improvement strategies. Such strategies should span the entire health behaviour decision-making process and incorporate multiple dimensions of support. Ultimately, the journey map may inform the development and future evaluation of stage-specific decision support, continuity across care transitions, and out-of-hospital management services for patients with T2DM who are repeatedly hospitalized.
Across the trajectory of health behaviour decision-making, patients with T2DM who experienced repeated hospitalisations showed stage-specific behavioural patterns, including experience-based care-seeking decisions marked by hesitation before admission and structured adherence under healthcare professional guidance during hospitalisation. The pre-admission hesitation may be attributed to the fact that repeated illness experiences do not necessarily translate into stable risk recognition and timely action. Before admission, patients may remain susceptible to delayed decision-making due to bounded rationality, previous experiences, and real-life circumstances.10,29 However, once admitted to hospital, the professionally led care structure reduces uncertainty while also making patients more likely to replace personal judgement with task execution and delegate final decisions to healthcare professionals based on limited participation.30 Subsequently, patients’ behaviours shifted from structured adherence during hospitalisation to fragile self-management during home management.
This transition may be attributed to the shift in care settings from hospital to home, where patients are required to assume greater responsibility for disease management and adapt to a changing role. Meanwhile, the accumulation of long-term management burdens and changes in social support networks may further undermine the stability of behavioural maintenance.31,32 Therefore, stage-specific and progressive interventions should be developed to address this behavioural transition trajectory. Specifically, risk alert education and guidance to facilitate timely healthcare-seeking decisions should be strengthened before admission;33 during hospitalisation, interventions should support patients’ transition from task-oriented adherence to active self-management;34,35 and during home management, continuous follow-up and feedback could be considered to support the sustained maintenance of health behaviours outside hospital settings.36
Throughout the trajectory of health behaviour decision-making, patients with T2DM who experienced repeated hospitalisations demonstrated stage-specific patterns of support needs, including seeking trusted decision anchors before admission, seeking structured empowerment and transition preparation during hospitalisation, and ultimately to seeking continuous and accessible out-of-hospital support during home management. These stage-specific patterns may be attributed to the fact that, when confronted with complex information and multiple treatment options before admission, patients’ unmet needs were not simply related to insufficient information, but rather to the lack of professional interpretation and decision-making guidance that could reduce uncertainty.37 Once treatment pathways became relatively clearer during hospitalisation, patients’ support needs further shifted towards understanding treatment plans, receiving specific management guidance, and preparing for transitions between hospital-based care and home management. This suggests that, at this stage, patients require clear, actionable, and continuous discharge support to facilitate a smooth transition from inpatient treatment to out-of-hospital management.38 However, when care transitions from hospital settings to the community, patients are required to independently undertake long-term management tasks within their everyday lives.
Deficiencies in the accessibility, responsiveness, and continuity of transitional care further intensify their need for ongoing support.39,40 Therefore, support systems could be optimised according to the evolving trajectory of patients’ needs. Specifically, before admission, patients could be provided with authoritative, clear, and interpretable decision support; during hospitalisation, greater emphasis should be placed on treatment explanation, specific management guidance, and preparation for transitions between hospital-based and home-based care; and during home management, continuous, convenient, and responsive out-of-hospital support mechanisms could be explored to address gaps in the accessibility and appropriateness of support.
Patients with T2DM who experienced repeated hospitalisations demonstrated stage-specific emotional experiences, including self-blame after initially neglecting the illness towards fragile stability within the hospital care context. These emotional experiences may be influenced by multiple factors, including illness risk perception, disease progression, and family and social support.41–44 In the early stages of the disease, insufficient awareness of illness severity may lead patients to experience regret and self-blame after disease deterioration or subsequent hospitalisation.45 As treatment progresses, symptoms improve, and disease management experience accumulates, patients may gradually regain a sense of control over their condition and develop more stable emotional regulation and acceptance of the illness.46 However, the lifelong nature of diabetes management and its inherent uncertainty may continue to trigger negative emotions, such as worry and fatigue, particularly in out-of-hospital settings.47 Therefore, emotional support could be integrated throughout the continuum of care for patients with T2DM who experience repeated hospitalisations. Through continuous emotional assessment, targeted psychological support, cultivation of adaptive coping strategies, and strengthening of family and social support networks, healthcare providers may facilitate patients’ emotional adjustment and long-term psychological adaptation.
The stage-specific pain points experienced by patients with T2DM who undergo repeated hospitalisations indicate that sustained health behaviour decision-making depends not only on the amount of disease-related knowledge patients possess, but also on whether they receive support that is aligned with the decision-making demands of each stage of care. In the early stage, patients’ limited ability to appraise disease-related information, together with information overload and inadequate health literacy, may lead to suboptimal decisions; the central difficulty lies in uncertainty surrounding the assessment of their condition and the choice of whether and where to seek care.48 During hospitalisation, readmission could provide an important opportunity for patients to re-examine their previous management experiences. Previous studies have suggested that when assessments of the reasons for readmission are conducted primarily from a hospital-centred perspective, without systematically integrating patients’ out-of-hospital experiences and the differing perspectives of patients, family members, and healthcare professionals, potentially modifiable gaps in management may be overlooked.49,50 This may prevent patients from translating the experience of readmission into a fuller understanding of long-term risks and stronger motivation for sustained disease management.
Additionally, some patients also expressed concerns about the perceived benefits and financial costs of TCM-related interventions, such as moxibustion. One possible explanation is that patients may weigh perceived treatment effects against financial investment when evaluating the value of these interventions, with limited perceived benefits potentially reducing their willingness to continue such treatments.51 During home management, although patients gradually recognise the importance of self-management, the demands of long-term management, together with family, work, and social pressures, may continually deplete their self-regulatory resources. In the absence of timely professional feedback and stable family support, behavioural intentions are less likely to translate into sustained action, and confidence in self-management may become increasingly fragile.52,53 Accordingly, the findings suggest potential intervention approaches tailored to each stage of care: before admission, risk clarification and support for information appraisal could be strengthened;37 during hospitalisation, multidisciplinary healthcare teams and patients could collaboratively review the factors contributing to readmission and prepare for long-term management, with nurses clarifying the purposes and realistic expectations of TCM-related nursing interventions, regularly assessing patients’ perceived benefits, and addressing their concerns to help them evaluate the value of these interventions;54 and after discharge, mechanisms for continuous professional feedback and family involvement could be explored, with telehealth and social support resources considered as potential means to minimise the disruption of long-term management by everyday circumstances.53
Limitations
This study has several limitations. Although the sample size was sufficient to provide the depth required for a longitudinal qualitative study, the findings arose from a specific healthcare and cultural context in mainland China, and the study was conducted at a single tertiary Grade A hospital of traditional Chinese medicine. Their transferability to other settings may therefore be limited.
In addition, the journey map of health behaviour decision-making was developed primarily from the patient perspective and did not incorporate the experiences of other stakeholders, such as caregivers and healthcare professionals. The map also mainly captured the shared trajectories of health behaviour decision-making among patients with T2DM who experienced repeated hospitalisations, with limited representation of heterogeneity across patient subgroups within the same stage. Future research could adopt cross-regional, multicentre designs involving different healthcare institutions and models of care, integrate the perspectives of multiple stakeholders, and use stratification approaches such as patient profiling to identify subgroup-specific decision-making pathways and support needs.
This may enhance the explanatory value of the journey map and improve the precision of intervention targeting. Mixed-methods designs could also be used to link journey-mapping findings with quantitative indicators, such as readmission, glycaemic control, and self-management, thereby supporting application across a wider range of settings.
Conclusions
This longitudinal qualitative study, using a patient journey mapping approach, provides important insight into the lived experiences of health behaviour decision-making among patients with T2DM who experience repeated hospitalisations. The resulting journey map offers a useful tool for elucidating the dynamic interactions among patients’ behaviours, needs, emotions, and pain points across different stages of care. By identifying critical disruptions in decision-making and patient-perceived gaps in care delivery at the study institution, this study provides a foundation for the future development and evaluation of stage-specific, patient-centred strategies for decision support and continuity of care.
Potential strategies for future development and evaluation include proactive risk education, reliable health information, and timely decision support to facilitate appropriate care-seeking before admission; nurse-led pre-discharge planning to address individual self-management needs and prepare patients for care transitions; and accessible telephone or digital follow-up could provide timely professional support during home management.