Insights into diabetes remission services: perspectives from general practitioners, family physicians and multidisciplinary teams Authors: Pichanun Mongkolsucharitkul, Preeyanan Chainarongloka, Rachata Walsri, Thiwat Sajjapanichkul, Sureeporn Pumeiam, Theerapat Thearachote, Puwadol Polpuak, Akapol Phisarn, Supachai Krobtrakulchai, Jatuphoom Neelasri, Krisada Hanbunjerd, Korrakot Weratean, Korapat Mayurasakorn Published: 2025 ABSTRACT AbstractObjectiveDiabetes remission has emerged as an achievable treatment goal, shifting the focus of care from increasing medication use to restoring metabolic health. While clinical trials show that remission is possible in controlled settings, evidence remains limited regarding its implementation in routine care within middle-income, rice-based dietary contexts. This study aims to explore healthcare provider experiences with implementing diabetes remission services in Thailand, focusing on dietary strategies, deprescription practices and patient management in routine care settings.DesignQualitative study using semi-structured, in-depth interviews, supplemented by structured questionnaires and programme documents.SettingThirteen healthcare facilities across six Thai regions and two national-level professional or policy organisations.Participants17 key informants purposively sampled for regional, institutional and professional diversity, including physicians, nurses, dietitians and national programme leaders. Data were collected and analysed iteratively until no new insights emerged. Thematic content analysis was conducted in QDA Miner Lite v3.0 with investigator triangulation.ResultFive major themes emerged: key strengths, success factors, nutritional approaches and lifestyle modification, implementation challenges, and development strategies. Multidisciplinary teamwork, personalised care plans and regular monitoring facilitated service delivery. Culturally adapted dietary strategies, such as low-carbohydrate Thai-style meals and intermittent fasting, were widely used. Challenges included unclear clinical guidelines, limited staffing and technological disparities. Medication deprescription varied across sites due to the absence of standardised protocols. Healthcare providers emphasised the need for community engagement and policy support to enable scale-up. Real-world implementation of diabetes remission services is feasible but challenged by systemic constraints and contextual variability. Flexible, culturally tailored approaches, empowered care teams and supportive policy frameworks are essential for sustainability.ConclusionThese findings provide practical insights for scaling remission programmes in other middle-income settings. Flexible, culturally tailored clinical pathways, empowered teams and supportive policy and financing are required to sustain outcomes and expand coverage.Keywords: Qualitative Research; Primary Health Care; Physicians, Family; Physicians, Primary Care; Diabetes Mellitus, Type 2 WHAT IS ALREADY KNOWN ON THIS TOPIC Lifestyle-based diabetes remission can be achieved in clinical trials. Evidence on real-world delivery in middle-income, rice-based dietary contexts is limited and healthcare provider perspectives are not well described. Lifestyle-based diabetes remission can be achieved in clinical trials. Evidence on real-world delivery in middle-income, rice-based dietary contexts is limited and healthcare provider perspectives are not well described. WHAT THIS STUDY ADDS Across diverse Thai settings, providers operationalise remission through multidisciplinary teamwork, personalised nutrition that includes Thai-style low-carbohydrate options, structured monitoring and flexible deprescribing. Barriers include cultural eating patterns, staffing constraints and digital access gaps. Across diverse Thai settings, providers operationalise remission through multidisciplinary teamwork, personalised nutrition that includes Thai-style low-carbohydrate options, structured monitoring and flexible deprescribing. Barriers include cultural eating patterns, staffing constraints and digital access gaps. HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY Findings inform scalable, context-adaptable pathways. Priorities include standardised yet flexible deprescribing guidance, workforce development for dietitians and nurses, and reimbursement for lifestyle services. Metrics for implementation evaluation in similar middle-income contexts are proposed. Findings inform scalable, context-adaptable pathways. Priorities include standardised yet flexible deprescribing guidance, workforce development for dietitians and nurses, and reimbursement for lifestyle services. Metrics for implementation evaluation in similar middle-income contexts are proposed. Introduction Diabetes mellitus is a global health epidemic and a leading cause of morbidity and mortality.1 Over 589 million adults are living with diabetes worldwide, a number projected to exceed 853 million by 2050.2 In Thailand, approximately 11.6% of adults (over 5 million people) are living with diabetes.3 Type 2 diabetes has long been seen as a chronic, progressive disease requiring lifelong pharmacotherapy. However, growing evidence shows remission is achievable in a significant subset of patients.4 5 Diabetes remission is defined as maintaining glycated haemoglobin (HbA1c)<6.5% for at least 3 months without glucose-lowering medications.4 Remission has been seen following intensive lifestyle interventions and bariatric surgery in selected cases,5 contributing to a paradigm shift in diabetes care. Intensive lifestyle modification and weight loss are central to achieving remission. The Look AHEAD trial, a large randomised controlled trial (RCT) of intensive lifestyle intervention in people with type 2 diabetes, reported a 1-year remission rate of 11.5%.6 In contrast, the primary care-based DiRECT cluster RCT achieved 46% remission at 12 months using a very low-calorie diet,7 36% of participants in the intervention group remained in remission at 2 years compared with 3% in the control group, with outcomes closely linked to sustained weight loss.8 Other dietary strategies, including low-carbohydrate Mediterranean diets, have also demonstrated higher remission rates and delayed medication use.9 Substantial weight loss through intensive dietary therapy can normalise glycaemic control in many patients. The physiological mechanisms behind remission are increasingly understood. Caloric restriction reduces hepatic fat and insulin resistance, improving fasting glucose and insulin levels.10 Continued weight loss lowers pancreatic fat, mitigating glucotoxicity and lipotoxicity, and partially restores β-cell function.10 Reducing carbohydrate intake lessens postprandial glycaemia and insulin demand, providing β-cell rest and enhancing insulin sensitivity.11 Thus, intensive lifestyle changes can reverse the metabolic defects of type 2 diabetes, not just manage glucose levels. Since remission excludes medication use, deprescribing usually starts with the intervention,4 requiring careful monitoring for safety while allowing lifestyle alone to maintain glycaemic control. Long-term remission remains challenging due to weight regain. In DiRECT, nearly half of the participants in remission at 1 year relapsed by year 2, mainly due to regaining weight.8 This underscores the need for ongoing dietary counselling, behavioural support and structured follow-up. Sustained engagement strategies, including regular check-ins or peer support, may help prevent relapse. In Asia, Thailand in particular, traditional diets are high in refined carbohydrates, such as white rice and noodles, and obesity manifests at lower BMI thresholds than in Western populations. Dietary programmes must align with local food practices. Thai authorities have begun integrating remission into national guidelines. In 2022, the Royal College of Family Physicians of Thailand released clinical guidance for achieving diabetes remission through intensive lifestyle modification,12 followed by the Department of Disease Control’s “Diabetes Remission Service” in 2023.13 These guidelines provide the national framework for eligibility, regular monitoring and deprescribing to minimise hypoglycaemia and related risks, but do not mandate a single dietary protocol. Services may implement several clinician-supervised modalities, such as low-carbohydrate diets, low- and very-low-calorie diets with or without meal replacement, plant-based diets adapted to Thai staples (eg, whole grains/unpolished rice) and time-restricted/intermittent fasting, with tailoring to patient preference, clinical risk, staffing and the local food environment.12 13 Total diet replacement is not required nationally. The DiRECT study serves as background evidence rather than a description of the Thai protocol. The 2023 national guideline formally recognised remission as a therapeutic goal,14 and the Diabetes Association of Thailand recently issued recommendations reinforcing remission in routine care,15 aligning with global efforts to make remission a standard treatment goal.16 Nonetheless, real-world feasibility in diverse Thai healthcare settings remains uncertain. Most evidence comes from trials in high-income countries, and little is known about how these interventions work in everyday practice in low- and middle-income settings. Provider perspectives are particularly lacking. Qualitative research can reveal context-specific barriers, facilitators and necessary adaptations. Previous qualitative studies in Western settings have shown that strong provider engagement and family support are critical for adherence to intensive diets,17 and participants in DiRECT described life-changing health benefits but also challenges such as hunger, social eating pressures and fear of disease recurrence.18 Moreover, a primary care cohort study in the United Kingdom found that greater clinician engagement and follow-up were strongly associated with drug-free remission, highlighting the role of supportive care relationships.19 Such findings illustrate the nuanced human factors that quantitative outcomes alone cannot capture. To address this gap, we conducted a qualitative study to examine how diabetes remission services are being implemented in Thailand. We explored healthcare providers’ experiences to identify factors supporting or hindering the scale-up of lifestyle-based remission services in a real-world, middle-income setting. Methods Study design and participants This qualitative study investigated real-world diabetes remission practices across Thailand, emphasising regional variation in dietary strategies, medication deprescription and patient management approaches. Participants included healthcare professionals directly involved in diabetes remission services, such as physicians, nurses, dietitians and representatives from professional medical associations. Participants were purposively sampled to ensure diversity across regions and healthcare settings. Two groups were recruited: (1) representatives from diabetes remission clinics who had experience in deprescription and lifestyle interventions, and (2) experts involved in guideline development or national programme implementation. Eligible participants were those actively involved in diabetes remission management and able to participate in online interviews. No exclusion criteria were applied; participation was voluntary, and withdrawal was allowed at any time. Of the 20 healthcare facilities and professional or policy organisations contacted by email or personal invitation and invited to individual interviews, 13 facilities and 2 organisations expressed interest in participating. This study was conducted following the Declaration of Helsinki. The study protocol was approved by the Institutional Review Board of the Faculty of Medicine Siriraj Hospital, Mahidol University (COA No. Si 802/2024). Informed consent was obtained from all participants. Data collection Data were collected between December 2024 and March 2025. The researcher who carried out the data collection had no professional or personal relationship with the participants. Interview guides and structured questionnaires were developed based on the preliminary review of Thai national guidelines, including clinical protocols by the Royal College of Family Physicians of Thailand and the Department of Disease Control. Data collection tools focused on deprescription processes, dietary strategies, patient follow-up systems and perceived barriers and enablers to service delivery. We used neutrally worded, open-ended prompts followed by systematic probes on barriers, failures, unintended consequences, and deprescribing challenges. The guide also included counterfactual probes to avoid celebratory narratives. Participants first completed a structured questionnaire covering clinical roles, practice patterns and institutional support. This was followed by a semi-structured Zoom interview (30–60 min) by the main author (female research assistant/dietitian, experienced interviewer), which was audio-recorded and transcribed verbatim. It was observed by two male and one female fourth-year medical students, who took notes to support transcription. No interviews were repeated. Supplementary documents, such as dietary protocols, clinic workflows and follow-up tools, were obtained. Data were collected and analysed iteratively until no further data yielded new insights. The researchers discussed and agreed when saturation had been reached. The transcripts were not returned to the participants for comment and/or correction due to timing constraints and to minimise burden on busy clinicians. To ensure interpretive rigour, two researchers independently coded all transcripts with consensus resolution, triangulated interviews with questionnaires and programme documents and monitored thematic saturation. Additionally, all participants reviewed the draft results and provided comments and factual corrections. Data analysis Thematic content analysis was performed using QDA Miner Lite V.3.0. Two researchers independently coded the data, with iterative discussions to resolve discrepancies. Key themes were synthesised and categorised systematically. Frequencies of recurring terms and concepts were tabulated. Special attention was paid to variations in dietary strategies, deprescription processes and patient engagement approaches across regions. Triangulation was achieved through the integration of multiple data sources, including interviews, questionnaires and programme documents, to enhance the credibility and robustness of the findings. Participant identifiers were anonymised throughout the analysis to ensure confidentiality. Patient and public involvement Patients and members of the public were not involved in the design, conduct, reporting or dissemination plans of this study. Our study specifically examined service delivery practices by healthcare professionals and policy stakeholders responsible for diabetes-remission services across Thailand; consequently, recruitment and data collection focused on clinicians, dietitians, nurses and professional/policy representatives rather than patients. We acknowledge this as a limitation, and the absence of patient/public input may have influenced topic prioritisation and interpretation. Results In-depth interviews were conducted with 15 representatives from 13 healthcare facilities, the delegate of the Division of Non-Communicable Diseases at the Department of Disease Control and the delegate of Thai Dietetic Association (table 1). Thematic analysis of the 17 interviews identified five major themes: (1) key strengths, (2) success factors, (3) nutritional approaches and lifestyle modification, (4) implementation challenges and (5) development strategies (figure 1; online supplemental table S1), with illustrative examples of remission service practices (online supplemental appendix 1-5). Table 1Participant characteristics and service models of diabetes remission programmes across healthcare organisations in ThailandRegion in Thailand: OrganisationParticipant ID: Role - Sex (Years of experience in diabetes management/ diabetes remission)Inclusion criteriaType of serviceCommunicationCentral: Private hospital 1P01: General practitioner - Male(2/2)HbA1c <9.0%, diagnosed with type 2 diabetes for less than 1–2 years, aged below 50 yearsNCD clinic*, non-intensive programme, individual careFace-to-face (30–40 min)Central: Private hospital 2P02: Internal medicine physician - Male (27/6)Newly diagnosed type 2 diabetes, any age; diabetes duration<30 yearsPrivate clinic, ILI†, individual careFace-to-face, online platforms (Line, Zoom)Central: Private hospital 3P03: Endocrinologist - Male (15/6)Case-by-case, FBS <80–130 mg/dL, HbA1c <7%, diabetes duration <5 yearsNCD clinic, individualised selection; intensive ILI with staged follow-upFace-to-face, online platforms (Line)Central: Primary care unit 1(in tertiary care hospital)P04: Family physician - Male (20/5)HbA1c <9.0%, or patient-motivated; no limitation on age or diabetes durationNCD clinic, ILI and non-intensive programme, both group and individual careFace-to-face, self-help group counsellingCentral: Tertiary care hospitalP05: Dietitian - Male (10/2)No strict criteria, but better outcomes observed in younger or newly diagnosed type 2 diabetes (<5–6 years)NCD clinic, ILI, individual care, structural diet programme on personalised nutritionFace-to-face, online platforms (Line)Central: Division of Non-Communicable Diseases, Department of Disease ControlP06: Physician - Male–––Central: Thai Dietetic AssociationP07: Dietitian - Female–––North: F1-level community hospital 1‡P08: Preventive medicine physician - Male (20/4)Adults (≥18 years) with type 2 diabetes, non-pregnant, eGFR≥45 mL/min/1.73 m²NCD clinic, ILI, both group and individual careFace-to-face, IT platform, online platforms (Line)North: M2-level community hospital 1§P09: Family physician - Male (13/1.5)Depends on the criteria specified for each service track or conditionNCD clinic and DM remission clinic, ILI with four categories: normal NCD clinic, admit hospital, Line Official Account and A-MED Home Ward¶, both group and individual careFace-to-face, online platforms (Line)Northeast: F1-level community hospital 2P10: Nutritionist - Female (20/2)Participation is voluntary. Most participants have a BMI ≥24 kg/m2 and HbA1c >7%.Diabetes remission clinic, ILI, both group and individual careFace-to-faceEast: Primary care unit 2(in tertiary care hospital)P11: Family physician - Male (7/3)Diagnosed with type 2 diabetes ≤5 years and BMI>25 kg/m2, final eligibility was considered at physician’s discretion.NCD clinic, ILI, both group and individual careFace-to-face, telemedicine, workshop, online platforms (Line)East: Primary care unit 3P12: Family physician - Female (6/1)Aged 30–65 years, newly diagnosed or with diabetes duration≤5 yearsNCD clinic (lifestyle modification clinic), non-intensive programme, individual careFace-to-faceWest: M2-level community hospital 2P13: Family physician - Male (20/2)Based on the 2023 standard guidelines, eligible patients are overweight or obese, have type 2 diabetes diagnosed within the past 5 years, demonstrate good glycaemic control (HbA1c ≤8%), and take no more than two medications, typically initial therapies such as glipizide and metformin.NCD clinic and diabetes remission clinic, non-intensive programme with three categories: smart, standard and remission service, both group and individual careFace-to-face, online platforms (Line)P14: Nurse - Female (1/1)South: M2-level community hospital 3P15: Family physician - Male (16/3)Type 2 diabetes for ≤5 years, with no severe complicationsDiabetes remission clinic, ILI, individual careFace-to-face, online platforms (Line)South: F2-level community hospital**P16: Preventive medicine physician - Male (16/5)No strict criteriaNCD clinic, ILI with three categories: intensive, smart and conventional, individual careFace-to-face, telemedicineP17: Nurse - Female (8/3)*Non-intensive programme was similar to general primary care practice and non-communicable disease care (NCD care).†Intensive lifestyle intervention (ILI) involves more time, more frequent intervention and interactive education, and a higher level of care and monitoring system.‡F1 (First Level 1) hospitals are community hospitals with 90 to 120 beds.§M2 (Middle Level 2) hospitals are community hospitals with more than 120 beds.¶A-MED Home Ward is a digital telehealth platform in Thailand designed to provide continuous home-based healthcare for patients who require inpatient-level care but are able to be monitored from home.**F2 (First Level 2) hospitals are community hospitals with 30 to 90 beds.eGFR, estimated glomerular filtration rate; HbA1c, glycated haemoglobin. Participant characteristics and service models of diabetes remission programmes across healthcare organisations in Thailand Non-intensive programme was similar to general primary care practice and non-communicable disease care (NCD care). Intensive lifestyle intervention (ILI) involves more time, more frequent intervention and interactive education, and a higher level of care and monitoring system. F1 (First Level 1) hospitals are community hospitals with 90 to 120 beds. M2 (Middle Level 2) hospitals are community hospitals with more than 120 beds. A-MED Home Ward is a digital telehealth platform in Thailand designed to provide continuous home-based healthcare for patients who require inpatient-level care but are able to be monitored from home. F2 (First Level 2) hospitals are community hospitals with 30 to 90 beds. eGFR, estimated glomerular filtration rate; HbA1c, glycated haemoglobin. Key themes and subthemes identified from healthcare practitioners’ experiences with the diabetes remission process. Theme 1: key strengths Subtheme: multidisciplinary team Remission success relied on multidisciplinary care involving physicians, nurses, nutritionists, pharmacists, physical therapists and community health volunteers, with team composition varying by setting (online supplemental table S2). Participants emphasised that team-based care aligned goals across disciplines, especially for nutrition therapy and planned deprescription. One participant stated, A strong, cohesive team is vital, where everyone shares the same understanding… Each member must know their role. – P08 Nutritionists and dietitians were described as pivotal for individualised meal planning, patient engagement, and effective communication: One of our strengths is the capable team of dietitians. They can personalise diets and communicate effectively to engage patients. This is supported by strong collaboration with other healthcare professionals. – P05 Nurses’ long-standing relationships with communities in primary care settings were repeatedly cited as an asset that translated trust into programme uptake and persistence: Having a respected nurse in the community is very helpful…Her strong connection through home visits and regular care has earned trust, making it easier to implement new programmes. – P09 Subtheme: personalised approach Most diabetes remission programmes applied eligibility criteria including HbA1c <8–9%, diabetes duration <5–6 years, age 30–65 and BMI ≥24 kg/m2; some also required preserved kidney function and absence of complications (table 1). A few considered patient motivation for lifestyle change. A key strength across programmes was the emphasis on personalised care, tailored to clinical status, behaviour and preferences. Services were delivered through individualised or hybrid group/individual formats, supported by multi-channel platforms, including face-to-face consultations, online messaging apps, video calls and telemedicine to support tailored follow-up and care planning. Longer consultations, typically 20 to 40 min, enabled goal setting and responsive counselling. One clinician described, A personalised, problem-based approach, where physicians assess each case individually, and patients join tailored sessions based on their specific needs, not a standard programme. – P16 Personalisation also incorporated patients’ daily experiences and preferences in treatment decisions. Patients are offered options and encouraged to choose what suits them best, with some able to try multiple methods to find what works for their condition. – P05 Subtheme: monitoring Monitoring was a core strength of remission services, with standard parameters, including body weight (BW), HbA1c, fasting blood glucose (FBG), postprandial glucose (PPG), blood pressure (BP) and lipid profiles (online supplemental table S3). More advanced facilities additionally incorporated body composition analysis through bioelectrical impedance analysis (BIA), fasting insulin, creatinine (Cr), C-peptide, HOMA-IR and dietary records including food feedback via mobile applications. These parameters served as essential indicators for evaluating patient progress, guiding medication deprescription and tailoring lifestyle interventions. Monitoring frequency ranged from every 1–4 weeks early on to every 2–3 months in stable cases. Digital tools such as glucometers, pedometers, smartwatches and platforms like LINE Official Account and Google Forms facilitated self-monitoring, accountability and timely feedback, though data were verified in clinic to ensure reliability: Digital data is downloaded and compared with written records…Verified data is used to calculate average glucose levels, with counselling based on both the readings and patient behaviour. – P17 Self-reported data may be unreliable. It is verified monthly against glucometer data. Patient data is sorted into individual spreadsheets for continuous, personalised monitoring. – P14 Subtheme: clear processes and policies Clear roles, defined workflows and visible leadership supported consistency and efficiency. Responsibilities are clearly divided, especially when it comes to policy. It’s important to set a firm goal and tell everyone exactly what needs to be done. – P13 Leadership that engaged directly in care reinforced adherence: The director not only sets the policy but also takes an active role in its implementation…directly involved in patient care and…clinical consultations… – P14 Theme 2: success factors Subtheme: patient motivation Intrinsic motivation was widely viewed as the engine of success. The key factor for success is that the patient must have the desire to improve whether it is to recover from the disease or to reduce their medication. – P01 Sustained motivation was supported through clear goals, motivational interviewing and SMART goals (Specific, Measurable, Achievable, Relevant, Time-bound) setting. Consistent follow-up is a key element in ensuring the sustainability of outcomes. – P03 Subtheme: patient self-learning and engagement According to study participants, active self-learning and engagement potentially predicted better adherence. Structured counselling, digital resources and peer support encouraged self-management and interpretation of personal data. Real-time glucose feedback strengthened awareness and self-efficacy: Technology that provides real-time data enables patients to monitor themselves, serving as a tool for self-directed learning. – P15 Patients often monitored more frequently early on, then less as they understood their own responses: Patients tend to check their blood sugar more often in the beginning as they learn how their body responds. But once they understand themselves better, they usually start checking less frequently. – P05 Subtheme: staff skill and communication Healthcare providers’ skills and communication were key to effective remission services. Multidisciplinary teams highlighted empathetic, culturally sensitive communication and soft skills such as active listening and emotional support: It involves…the ability to listen…design personalised care plans…offer emotional support…When a doctor says, I’m here to listen…focused on you today, it can make a big difference. These are soft skills like empathy and good communication. – P16 Theme 3: nutritional approaches and lifestyle modification Nutritional and lifestyle interventions were culturally adapted based on patient needs and available resources (table 2, online supplemental table S4). Low-carbohydrate diets predominated, often paired with higher fibre and context-appropriate substitutions, such as brown rice. In areas where sticky rice is normative, satiety expectations complicated reduction strategies. Sequential modification shifted meal composition toward protein and fibre to reduce rice quantity: Table 2Nutritional approaches and lifestyle modificationRegion in Thailand: OrganisationLifestyle modification programmeNutritional approachesPhysical activity and exerciseCentral: Private hospital 130–40 min session, with dietary and exercise recommendations and tools providedPartial meal replacement (40%), high fibre diets (20–30%), plate model* (20%), low fat diet (10–15%), intermittent fasting (5–10%)Limited time: Increase daily steps. Walk 30 min/session, 150 min/week.Available time: Do cardio using heart rate zones (with monitor).Cardio established: Add weight training to build muscle mass.Central: Private hospital 2Dietary and exercise recommendations, sunlight exposure, grounding and sleep adviceVery low carbohydrate/low carbohydrate diet (100%), carnivore diet, intermittent fasting (18–6, 100%), one meal a day, alternate day fastingAerobic exercise (30%) for 20 to 40 min and 5 to 7 times/weekResistance training (70%) for 30 to 60 min and 2 to 3 times/weekHigh-intensity interval training for 20 to 30 min and 2 to 3 times/weekCentral: Private hospital 3Behaviour modification programme, recommended by dietitian or certified diabetes educatorCaloric restriction diet (100%), partial meal replacement (100%)Wellness consultant for 175 min/weekCentral: Primary care unit 1(in tertiary care hospital)20–30 min session individually and in group with mainly dietary modification and increased physical activity when possibleBrown rice, low carbohydrate and high protein, low saturated fat diets (60%), time-restricted eating (60%), plate model (20%), very carbohydrate diet (10%), avoid sweetsAerobic/resistance training, walkCentral: Tertiary care hospital30–40 min session, with dietary and exercise recommendationPlate model (100%), high fibre diet (60%), caloric restriction diet (60%), low carbohydrate diet (<130 grams of carbohydrate, 40%), intermittent fasting (16:8, 18:6, 40%), partial meal replacement (15%)Aerobic (walking/jogging/cycling based on patient preference) for 30 to 60 min/day and 3 to 5 times/weekResistance training (weightlifting/squad) for 15 to 30 min/day and 2 to 3 times/weekNorth: F1-level community hospital 1†60 min of group activity, 4 sessions, with dietary and exercise recommendations and tools provided (glucometer and strips, digital platforms)Low carbohydrate diet (<120 grams of carbohydrate, 100%), high fibre diet (100%), plate model (2:1:1) and intermittent fasting (16:8, 12:12, 100%)Aerobic (walk/run/bike) for 30 to 60 min and 3 to 5 times/weekResistance training (stretching/squad) for 30 min and three times/weekNorth: M2-level community hospital 1‡3 hours of group activity, every 3 months, with dietary recommendations and sharing experiencesVery low carbohydrate/low carbohydrate diet (100%), black coffeeEncourage increased physical activity but not strictNortheast: F1-level community hospital 2Carb counting with dietary and exercise recommendations at diabetes remission training centre (12 times/year)High fibre diet, low carbohydrate and high proteinAerobic dance (older adults) for 30 min and 3 to 4 times/weekBreathing technique for allWalking (adults)East: Primary care unit 2(in tertiary care hospital)2 hour group workshops (50 participants/group), 3 sessions, with dietary and exercise recommendations and tools provided (glucometer and strips)Plate model (100%) since this is per se low caloric and low carbohydrate ratiosWalk (up to 10 000 steps), 3 to 5 times/weekEast: Primary care unit 315 min session, with dietary and exercise recommendationsLow carbohydrate and high protein, intermittent fastingStrengthening exercises include patient-friendly movements (eg, wall sit, full plank, half leg raises, group brisk walk). Focus on muscle-building activities.West: M2-level community hospital 230–40 min session, with dietary and exercise recommendations and tools provided such as a glucometer (total 15 glucometers in the centre)Low carbohydrate diet (70%), avoid sweets, brown rice, high fibre diet (80%), plate model (60%), Mediterranean diet (20%)Aerobic (jogging/cycling on patient preference) for 30 min/day and 3 to 5 times/weekResistance training (weightlifting/squad) for 30 minutes/day and 3 to 5 times/weekStretching exercises for 30 min/day and 3 to 5 times/weekFlexible exercises for 1 hour/day and 3 to 5 times/weekSouth: M2-level community hospital 330–40 min session, with dietary and exercise recommendations and tools provided such as glucometerBrown rice and high fibre diet (100%), plate model (100%), intermittent fasting (16:8, 20%), low-salt fish sauce, avoid sweets and 3 in 1 coffee mixWalking (up to 10 000 steps)South: F2-level community hospital§20–30 min session, with dietary and exercise recommendations and tools provided such as glucometer and pedometerBrown rice, high fibre diet and plate model (80–90%), partial meal replacement (<10%)Walking, strength and balance exercises to increase physical activities*plate model: a dietary guide recommending half a plate of vegetables, a quarter of lean protein and a quarter of whole grains or complex carbohydrates.†F1 (First Level 1) hospitals are community hospitals with 90 to 120 beds.‡M2 (Middle Level 2) hospitals are community hospitals with more than 120 beds.§F2 (First Level 2) hospitals are community hospitals with 30 to 90 beds. Nutritional approaches and lifestyle modification plate model: a dietary guide recommending half a plate of vegetables, a quarter of lean protein and a quarter of whole grains or complex carbohydrates. F1 (First Level 1) hospitals are community hospitals with 90 to 120 beds. M2 (Middle Level 2) hospitals are community hospitals with more than 120 beds. F2 (First Level 2) hospitals are community hospitals with 30 to 90 beds. Changing from sticky rice to regular rice is hard because they’ve eaten it since childhood, but we encourage them to try regular rice and see how it goes. If some succeed, they can gradually adjust. – P08 The plate model was widely implemented as a guide emphasising balanced proportions of vegetables, lean protein and whole grains. Intermittent fasting, including protocols such as 16:8 and 18:6, was also integrated into many programmes. However, non-conventional diets like ketogenic or intermittent fasting were approached with caution. One expert stressed individualised assessment, professional oversight and time limits, noting no single diet fits all. Extreme dietary patterns were conditionally accepted with ongoing monitoring. Patients favouring ketogenic diets were allowed to try them short-term, with transition plans in place. Health monitoring ensured safety, especially with restrictive diets, while prolonged fasting was discouraged due to health risks. Participants avoided blanket endorsements of any one pattern: We don’t believe keto is entirely bad…Likewise, we don’t think IF is entirely good…we remain aware and open to all views. – P06 If a patient strongly believes in keto, we advise caution and limit its use to a short period. Once results are seen, we guide them to gradually adjust their plan. – P07 To support adherence, many facilities sourced and made available for purchase healthier options, eg, brown rice, black coffee and low-sodium fish sauce, at the clinics, making them readily accessible to patients. However, the limited availability of nutritious options in some communities remained challenging. Misalignment between diabetes remission goals and the local food environment poses a barrier, particularly in communities where unhealthy food options are prevalent. “We want this community to achieve DM remission, but there are no healthy foods…just unhealthy options everywhere. – P07 Structural factors, such as supportive policies and community infrastructure, were seen as critical for long-term success. In addition, the expert pointed to the value of preventive health investments, suggesting that policy frameworks should prioritise prevention not only for diabetes but for chronic diseases more broadly. The economic and health benefits of focusing on prevention were underscored as a key rationale for enhancing community-level resources. The participant argued that prevention-oriented investments would yield broader health and economic returns. Physical activity recommendations ranged from daily step goals to structured aerobic, resistance or high-intensity interval training (HIIT). Several programmes tailored exercise advice based on patient availability, physical capacity and the use of monitoring tools. One expert noted that exercise helps maintain muscle mass and regulate energy. Coordination between diet and physical activity was emphasised, with dietitians ensuring patients were nutritionally prepared for safe exercise. Exercise might not have a large effect…but…it boosts your metabolic rate…I think exercise is important. – P06 We also ensure blood sugar is checked beforehand because if it’s too high beyond the set threshold, patients shouldn’t exercise yet. – P07 Theme 4: implementation challenges Subtheme: dietary challenges Implementing dietary interventions faced challenges rooted in cultural habits and entrenched behaviours across diverse settings. Standard models like the Plate Model were difficult to apply in real life, especially in urban areas where reliance on commercial meals limited vegetable intake. When relying on purchased meals, it becomes challenging to achieve the recommended portion. – P01 Cultural food practices in southern Thailand, such as communal meals at special traditional ceremonies, also conflicted with dietary guidelines. Some funerals last up to seven days, with food served throughout. Eating is strongly tied to social and community events in southern Thailand. – P16 Additionally, high sugar consumption was noted as a prevalent issue, particularly due to cultural preferences for sweet foods. Participants reported that this habitual sugar intake posed a major obstacle to dietary change. However, some success was noted in encouraging gradual reduction, especially through focused behaviour modification strategies. People are used to eating sweet foods…it becomes a habit that’s hard to break…we encourage reducing sugar in coffee or avoiding sugary drinks, especially soft drinks. If they follow this for two to three weeks, many can adapt and maintain the change. – P13 Subtheme: unclear clinical guidelines Variability in treatment protocols, patient management and deprescribing criteria across settings has led to inconsistent clinical decisions, reflecting site-level choices within the nationally permitted modalities. Although remission is now a recognised goal, the lack of clear, widely accepted guidelines creates challenges. There is currently no clear protocol for managing diabetes remission…Although guidelines mention options…they lack detailed instructions on when to use or switch between these approaches. – P05 This affects follow-up schedules, dietary choices and individualised care, forcing providers to rely on personal judgement or local protocols. Rigid guidelines, where available, may not suit all settings, especially those with limited resources. It is not feasible to apply every step exactly…It is necessary to adapt and tailor the guidelines to suit the specific context and needs of our own setting. – P14 Participants emphasised the need for flexible but clear clinical pathways adaptable to diverse contexts. The absence of standardised guidelines extends to medication deprescription strategies, particularly concerning the reduction or cessation of antihyperglycaemic agents during the course of diabetes remission. Currently, there are no universally accepted protocols that dictate the precise order or criteria for deprescribing medications, which has led to variability among healthcare institutions in practice. While most prioritise stopping insulin and sulfonylureas first, approaches still depend on patient response, monitoring data and clinician judgement. Deprescription often considers weight loss, FBG and HbA1c, but execution varies due to differences in resources, clinical expertise and patient profiles (table 3, online supplemental table S5). These inconsistencies highlight the need for clearer, adaptable deprescription guidelines. Table 3Medication deprescription and target blood glucose levelsRegion in ThailandOrganisation*Stepwise deprescription orderTarget parameters for further deprescribingCentralPrivate hospital 1Insulin → SFU → TZD → MFM → DPP-4/SGLT2TZD may be retained in patients with hepatic steatosis.HbA1c <6.5–7.0%Monitor for side effects/safety and BW loss. If BW reduces by 5–10%, proceed to the next step.Private hospital 2Insulin → SFU → SGLT2 → TZD → DPP-4 → MFM with blood glucose, BW and WC monitoringHbA1c <7.5%, or FBG <150 mg/dLHbA1c <6.5% is the ultimate blood glucose targetPrivate hospital 3Insulin → SFU → MFM → DPP-4, considering case by caseHbA1c <7%, or FBG 80–130 mg/dLPrimary care unit 1(in tertiary care hospital)Insulin → SFU → TZD → MFM → DPP-4/SGLT2 Monitor for side effects and BW. For patients on high doses of SFU or MFM, consider reducing to half dose.HbA1c<7.5–8.0%, or if HbA1c is better or comparable to the previous one or patients have hypoglycaemic symptomsHbA1c<6.5% or 1 hour PPG<155 mg/dL are the ultimate blood glucose targetsNorthF1-level community hospital 1†Insulin → SFU → TZD → DPP-4 → MFMSGLT2 inhibitors are not used for people with low carb/IF intervention. Monitor for side effects/safety and BW.HbA1c<7%, or FBG<126 mg/dL2 hour PPG<140 mg/dL, continuously 2–4 weeks, SMBG is better, BW reduces or patients have hypoglycaemic symptomsM2-level community hospital 1‡Insulin → SFU → TZD → MFMDiscontinue insulin with successful very low-carb dietSFU: stop if 2 hour PPG<110 mg/dL; reduce dose if 2 hour PPG is between 110–140 mg/dLTZD/MFM: stop if HbA1c ≤5.5%; reduce dose if HbA1c is between 5.5 and 6.5%EastPrimary care Unit 2(in tertiary care hospital)Insulin → SFU → TZD → MFMMonitor for side effects/safety and BW. If BW decreases by 5–10%, continue deprescribing. Retain TZDs if the patient has steatotic liver disease. In the case of patients having steatotic liver disease, TZD might be retained.HbA1c<7.5%, or FBG<150 mg/dL or if HbA1c is better or comparable to the previous one or patients have hypoglycaemic symptomsHbA1c<6.5% or FBG<110 mg/dL are the ultimate blood glucose targetsPrimary care unit 3If the patient adopts dietary changes, follow-up in 1 month to reassess FBS. If FBG>140 mg/dL, maintain current medication. If FBG<140 mg/dL, stop hypoglycaemic agents (SFU). Adjust medication based on follow-up and adherence.FBG<140 mg/dLWestM2-level community hospital 2Insulin → SFU → TZD → MFM → DPP-4/SGLT2HbA1c<6.5%SouthM2-level community hospital 3SFU → Insulin→ TZD → MFM with monitoring every 1–2 weeksHbA1c<6–6.5%, or FBG<110–120 mg/dLF2-level community hospital§SFU → Insulin→ TZD with blood glucose monitoringMean PPG<120, continuously 1–2 weeksHbA1c<6.5% is the ultimate blood glucose target*Data shown includes only healthcare facilities where interviews were conducted with physicians who were authorised to prescribe medication.†F1 (First Level 1) hospitals are community hospitals with 90 to 120 beds.‡M2 (Middle Level 2) hospitals are community hospitals with more than 120 beds.§F2 (First Level 2) hospitals are community hospitals with 30 to 90 beds.BW, body weight; DPP-4, dipeptidyl peptidase-4 inhibitors; FBG, fasting blood glucose; HbA1c, glycated haemoglobin (haemoglobin A1c); IF, intermittent fasting; MFM, metformin; PPG, postprandial glucose; SFU, sulfonylureas; SGLT2, sodium-glucose cotransporter two inhibitors; SMBG, self-monitoring of blood glucose; TZD, thiazolidinediones; WC, waist circumference. Medication deprescription and target blood glucose levels Data shown includes only healthcare facilities where interviews were conducted with physicians who were authorised to prescribe medication. F1 (First Level 1) hospitals are community hospitals with 90 to 120 beds. M2 (Middle Level 2) hospitals are community hospitals with more than 120 beds. F2 (First Level 2) hospitals are community hospitals with 30 to 90 beds. BW, body weight; DPP-4, dipeptidyl peptidase-4 inhibitors; FBG, fasting blood glucose; HbA1c, glycated haemoglobin (haemoglobin A1c); IF, intermittent fasting; MFM, metformin; PPG, postprandial glucose; SFU, sulfonylureas; SGLT2, sodium-glucose cotransporter two inhibitors; SMBG, self-monitoring of blood glucose; TZD, thiazolidinediones; WC, waist circumference. Subtheme: limited staff Human resource constraints were a major challenge to implementing and scaling diabetes remission services. Staffing shortages made it difficult to adapt or expand programmes. The dual modality of service delivery, online and on-site, was also affected by staffing limitations. While online services offer flexibility for patients, they shift a disproportionate amount of operational responsibility onto healthcare providers. Online services place a heavy workload on healthcare staff, as they must handle registration, data entry, interviews, consultations, and medication delivery…high patient volume makes them difficult. – P16 Furthermore, reliance on a small number of key personnel created vulnerabilities in maintaining consistent service quality. For instance, clinics that depended on a limited number of dietitians resulted in increased workload and disrupted continuity of care. Our hospital has two dietitians, and they divide their time between several clinics. One of them mainly takes care of this clinic. But if she takes time off or is unavailable, it becomes a problem for us and can affect the continuity of care. – P14 Subtheme: monitoring and follow-up Monitoring and follow-up were major challenges due to unequal access to technology and limited digital literacy. These challenges hindered continuous care and self-monitoring, often leading to dropouts and relapses. Some common issues we often face are, first, patients relapsing…loss to follow-up. – P08 Online services were often impractical, especially for groups like middle-aged housewives unfamiliar with digital tools. Most of the participants are housewives aged 40 to 50 years and older. This population generally does not engage with online platforms. – P10 Low education and poor digital access further limited participation, and the high cost of monitoring devices added financial strain. Subtheme: patient attitudes and lifestyles Patient habits, cultural norms and work demands were major challenges to remission efforts. Irregular schedules among vendors, shift work and agricultural routines often conflicted with dietary and medication adherence. It is considerably more challenging for market vendors. Many of them travel with produce trucks to sell goods at various markets in Bangkok. Their schedules are often unpredictable, which makes it harder for them to stick to regular meal times or take their medication as planned. – P14 Agricultural routines contributed to breakfast skipping and large evening meals shared with family, socially meaningful but metabolically challenging: About 50–60% struggle with blood sugar control due to habits like skipping breakfast or waiting to eat with family. – P10 Cultural preferences, like sticky rice consumption, also hindered dietary changes. Additionally, patient beliefs and emotional responses, such as stress from self-monitoring or peer comparisons, often led to disengagement. Providers emphasised that altering mindsets was often more difficult than delivering clinical care. Theme 5: development strategies Subtheme: scaling up implementation Advancing remission services requires scaling up through standardised, culturally adaptive protocols for screening and follow-up. Participants called for stronger interdisciplinary coordination and recognised system capacity as a critical factor. Promoting community ownership was seen to shift from provider-led to patient-led care, normalising remission as a shared goal. Scaling up is crucial…larger expansion needs proper monitoring and a dedicated team. Technology, such as AI tools, could also support this effort. – P08 …to establish a community of individuals in diabetes remission and communicate that achieving remission is both attainable and ordinary… – P09 Support for lifestyle medicine training and policy alignment was emphasised, along with expanding prevention clinics and integrating services into primary care for long-term sustainability. Subtheme: enhanced roles of the multidisciplinary team Participants emphasised the need for well-structured teams with expertise in nutrition, lifestyle medicine and patient education. Expanding teams to include dietitians, trained nurses and pharmacists was seen as essential. Enhancing nurses’ technical skills was viewed as key to improving monitoring and personalised care. Effective team coordination and cross-training were highlighted, while physician availability and capacity remained ongoing challenges. It would be beneficial for diabetes remission clinics to have a more comprehensive multidisciplinary team. For example, including nurses who serve as educators or dietitians could provide valuable support in this area. – P01 Subtheme: increased availability of equipment Expanding diabetes remission services requires adequate equipment and patient-centred educational tools. Participants emphasised that as patient numbers grow, it becomes imperative to scale up both clinical and supportive resources, particularly in terms of self-monitoring devices and personalised educational materials. Blood glucose monitors were identified as a key need for effective self-management. Additionally, facilities and resources to support dietary management, including adaptable meal plans and accessible education, were seen as crucial for lifestyle modification. One key concern is ensuring enough blood glucose monitors…educational booklets… Ideally, each patient should have a personal handbook with practical tips. – P11 Subtheme: technology-assisted monitoring Real-time tracking and digital systems were seen as vital for supporting both clinical outcomes and patient motivation. Participants emphasised the need for automated data collection and analysis to reduce provider workload. Advanced monitoring helps detect early signs of relapse or complications, such as hypoglycaemia, enabling timely intervention. Beyond technological infrastructure, participants highlighted the importance of establishing robust follow-up mechanisms. Monitoring should not cease on achieving initial remission milestones; rather, it requires sustained efforts over extended periods, typically spanning 1 to 2 years. We would likely need more funding…glucose monitors or systems that can collect and process data automatically. – P14 …focused on advancing technologies for detecting early warning signs of patient relapse…to improve the early detection of medical emergencies arising from treatment-related complications, specifically acute conditions such as hypoglycaemia… – P16 Expert insights on policy framework for implementing diabetes remission Experts described Thailand’s policy framework as flexible and locally adaptive, aiming to embed lifestyle interventions into routine care. Providers are encouraged to tailor approaches to culture and context: The Department has never said that everyone must follow the same approach…We want people to learn and adapt the approach to fit their own local context. – P06 Personalised care is central to this framework, with interventions tailored to individual preferences and lifestyles and aligned with the nationally permitted modalities. One example of this personalised approach in development is the Siam Diet, a dietary pattern tailored specifically for Thai people. Experts are working to adapt principles from well-known healthy diets, such as the Mediterranean or DASH (Dietary Approaches to Stop Hypertension) diets, using local Thai ingredients and cooking styles. The Dietitians Association of Thailand, along with other professional groups, is promoting the Siam Diet, a healthy eating pattern adapted for Thai people…Many healthy elements, such as vegetables and fish, are already common in Thai cuisine, and each region offers unique, healthful dishes. The main strength of the Siam Diet is its cultural fit. Thai people are familiar with these foods, making it easier to follow and enjoy every day. This makes the Siam Diet more sustainable and suitable for long-term use. – P07 However, experts noted the absence of clear deprescription guidelines and patient selection criteria, which hinders consistent clinical practice. Challenges include the high cost and limited availability of healthy foods, particularly in low-income areas, making adherence difficult. A key policy goal is to enable reimbursement for lifestyle services, offset by savings from reduced medication use. Limited staffing and lack of reimbursement for roles like dietitians also restrict implementation. Experts also stressed the value of biofeedback tools like continuous glucose monitors to support sustained engagement, though access remains limited. Overall, remission is viewed as a gradual process requiring long-term, ongoing support. Discussion Healthcare providers identified several key strengths supporting diabetes remission in Thailand. Multidisciplinary teamwork, including physicians, nurses, dietitians, etc., was central to coordinated, patient-centred care, consistent with global recommendations.20 Personalised care plans, regular monitoring, eg, frequent HbA1c checks and weight tracking, and clear clinical protocols enhanced provider confidence, reflecting strategies used in the DiRECT trial.21 Patient motivation and engagement were viewed as critical; informed and proactive individuals achieved better outcomes, supporting prior findings on self-efficacy.22 Conversely, low health literacy and passive attitudes limited success. Remission was used as a motivational concept to promote behaviour change.16 Effective communication and team alignment were essential for consistent care delivery.22 Dietary and physical activity strategies were culturally adapted. Rather than strict meal replacement regimens, providers promoted culturally acceptable dietary modifications to improve adherence.23 Intermittent fasting, supported by a Chinese trial showing up to 44% remission at 12 months,24 was used alongside culturally appropriate physical activity.7 Our participants recognised that helping patients find enjoyable, culturally acceptable ways to be active, eg, group exercises or daily lifestyle activities, was particularly important in low-activity communities. These findings support the need for flexible, context-specific interventions. Emerging studies in South Asian populations highlight how remission diets are delivered matters for acceptability. In public and patient involvement activities with South Asians in the UK, total diet replacement was viewed as unacceptable, whereas a culturally tailored, food-based low-energy diet was preferred to support adherence.25 In contrast, the STANDby feasibility trial in UK-based South Asians suggested that, when offered with structured support, total diet replacement can be acceptable for many and deliver mechanistically plausible benefits with substantial weight loss and reduced liver fat.26 Our provider accounts indicate that partial meal replacement combined with culturally adapted real-food meals is used in some Thailand settings, with no sites implementing full total diet replacement. Teams preferentially use low-carbohydrate diets as the foundational dietary strategy rather than complete formula-based replacement. These findings are consistent with the South Asian literature that raises acceptability concerns about total diet replacement. For Thai service design, our results support a dual-pathway approach that prioritises a culturally tailored low-carbohydrate diet with flexible partial meal replacement. Long-term adherence was a key challenge, particularly due to hunger, cravings and social pressures similar to issues seen in clinical trials.21 Flexible food choices supported ongoing engagement. The absence of standardised guidelines created uncertainty in follow-up care, consistent with reports from Scotland.27 Sustainability was also a concern, as weight regain and relapse are common without long-term support, reflecting the chronic nature of type 2 diabetes.16 Providers cited system-level constraints, including shortages of nurses and dietitians and limited time for follow-up, which are magnified in resource-limited settings.28 Healthcare providers faced difficulty balancing remission care with routine duties, leading to service inconsistencies. Socioeconomic barriers contributed to missed appointments, dropout and delayed relapse detection. To scale programmes effectively, providers recommended gradual expansion to more primary care units, supported by training and resource allocation. Fidelity and outcome monitoring were emphasised as essential.29 Task-shifting was highlighted as a promising strategy; empowering nurses, dietitians and community health workers could alleviate physician burden and improve reach, as shown in low- and middle-income countries.30 Many providers saw an opportunity to formally empower nurses and dietitians to take on greater roles in patient education, dietary counselling, and follow-up, which could alleviate physician workload and extend the reach of the programme. Additionally, engaging community health volunteers for patient outreach or peer support groups could help address adherence and follow-up challenges at the community level. Technology such as glucometers and mobile health apps was helpful, though cost and digital literacy were noted concerns. Telemedicine and LINE messaging, used during COVID-19, may support hybrid models to improve access, especially in rural areas. Although our data address diabetes remission services, several core components identified here, such as structured lifestyle support, diet-first pathways, safety-oriented deprescribing protocols and proactive follow-up, are plausibly adaptable upstream to high-risk or pre-diabetes populations. Contemporary commentaries propose pre-diabetes remission as a prevention goal, with operational criteria and signals that achieving remission after lifestyle-induced weight loss can substantially reduce subsequent type 2 diabetes risk, compared with weight loss targets alone.31 One participating site in our study plans to extend its model to individuals with pre-diabetes, underscoring service-design relevance; rigorous prospective evaluation will be required before routine adoption. For adaptation, services would need: (1) clear eligibility/readiness criteria beyond HbA1c alone; (2) stepped-care, lower-intensity protocols scalable in primary care; (3) monitoring/safety thresholds tailored to people not using glucose-lowering medications; (4) consensus data standards to define and track pre-diabetes remission and reversion and (5) reimbursement and quality indicators aligned with prevention goals. Policy support is essential. Participants called for remission to be formally recognised as a clinical goal, with standardised dietary protocols, eligibility criteria and follow-up schedules. Integrating remission into the universal coverage scheme is vital for sustainability. Strengths and weaknesses of the study Strengths include diverse sites across regions and facility levels, triangulation of interviews, questionnaires and programme documents, and transparent analysis with dual coding and consensus. Perspectives were limited to healthcare providers and policy stakeholders; the absence of patient interviews may bias interpretations toward service-delivery viewpoints and reduce insight into lived-experience barriers, eg, stigma, food insecurity and caregiving burdens. Some clinics collect patient-reported information, but these data were not consented for qualitative analysis and were not included. Provider narratives may be affected by social desirability or positive framing. Qualitative themes were not linked to objective outcomes, eg, HbA1c change, remission/relapse and medication-free duration because harmonised patient-level data were unavailable across sites. Finally, our inquiry focused on diabetes remission. We did not evaluate the emerging concept of ‘pre-diabetes remission’, for which definitions and evidence are still evolving. Future research Future work should integrate patient perspectives using a mixed-methods approach that purposively samples across socio-economic strata and programme completion status to assess acceptability, feasibility and equity. An explanatory sequential design should adopt a minimal common outcome, eg, HbA1c at 6 and 12 months, remission criteria, 12 month relapse and medication-free duration, with harmonised site reporting and routine data linkage. Quantitative analyses should test whether practices identified here, eg, protocolised deprescribing, food-environment modifications and multidisciplinary follow-up, predict superior outcomes, followed by qualitative inquiry to explain site-level differences. Evaluation of long-term outcomes and cost-effectiveness will guide scalable implementation. Conclusion This qualitative study reveals that diabetes remission services in Thailand are feasible but require nuanced, contextually adaptive implementation strategies. Our findings demonstrate that successful remission programmes depend on multidisciplinary teamwork, personalised care approaches and culturally tailored nutritional interventions. While healthcare providers recognise remission as a viable therapeutic goal, significant challenges persist, including unclear clinical guidelines, limited staffing resources and technological disparities. The study highlights the critical importance of flexible clinical pathways that can accommodate diverse healthcare settings and patient needs. Effective strategies include empowering multidisciplinary teams, leveraging digital monitoring tools and developing culturally sensitive dietary modifications. Key success factors encompass patient motivation, comprehensive self-learning approaches and robust communication between healthcare providers and patients. The research underscores that sustainable diabetes remission in middle-income settings requires not only clinical interventions but also supportive policy frameworks, workforce development and community engagement. Future implementation efforts should prioritise standardised yet adaptable protocols, invest in healthcare professional training and create the infrastructure that supports long-term patient follow-up and lifestyle modification. These insights provide a foundational roadmap for scaling diabetes remission services in resource-constrained environments. Supplementary material Acknowledgements The authors would like to express their sincere gratitude to the General Practitioners/Family Physicians Association of Thailand for their valuable academic consultation and expert guidance, which greatly contributed to the development of this study. The authors sincerely thank Sakditat Ittiphisit (Siriraj H Solutions), Thanasak Yimkerd (Private clinic), Jirathasri Suwanbumroongchai (Regional Health Promotion Center 6 Chonburi), Pichit Sugsabye (Bannasan hospital), Samitti Chotsriluecha (Thai Dietetic Association), Supranee Promsukhan (Thayang hospital), Chalisa Jantoy (Bantakhun hospital), Sumitra Jaisue (Nong Bun Mak hospital) and Ekgaluck Wanothayaroj (Vimut-Theptarin hospital) for generously sharing their time, insights and experiences in diabetes remission process. Their contributions were critical to the success of this study Footnotes Funding: The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors. Provenance and peer review: Not commissioned; externally peer reviewed. Patient consent for publication: Not applicable. Ethics approval: This study involves human participants and was conducted following the Declaration of Helsinki. The study protocol was approved by the Institutional Review Board of the Faculty of Medicine Siriraj Hospital, Mahidol University (COA No. Si 802/2024), dated 10 October 2024. Participants who agreed to participate in the study received information about the purpose and procedure of the study and the data collection process, including the video recording of the interview and the confidentiality of participants’ information. Participants gave informed consent to participate in the study before taking part and for their data to be published. In addition, the interviewer requested additional verbal consent from each participant before the interview was recorded. Data availability free text: Data are not publicly available due to privacy or ethical restrictions. Requests for access to the questionnaires or study materials can be directed to the corresponding author (korapat.may@mahidol.ac.th). Patient and public involvement: Patients and/or the public were not involved in the design, or conduct, or reporting, or dissemination plans of this research. Data availability statement Data are available upon reasonable request. References Associated Data Supplementary Materials Data Availability Statement Data are available upon reasonable request.