Global Landscape of Telehealth Interventions for Type 2 Diabetes Self-Care: A Scoping Review of Nursing Roles, Care Models, and Outcomes

Abstract

Background

Type 2 diabetes mellitus (T2DM) requires ongoing self-care support to minimize preventable complications. Telehealth interventions have the potential to extend diabetes self-management education and support beyond conventional face-to-face care. However, the specific contributions of nursing roles, conceptual frameworks, and digital care architectures have not been comprehensively synthesized.

Objective

This scoping review aimed to map global evidence on telehealth interventions for T2DM self-care.

Methods

The review was conducted in accordance with Arksey and O’Malley’s framework, JBI scoping review guidance, and the PRISMA Extension for Scoping Reviews. Electronic databases including PubMed/MEDLINE, EBSCOhost, Scopus, Taylor AND Francis Online, ScienceDirect, and Sage Journals, were systematically searched for empirical studies published between January 1, 2021, to June 26, 2026, examining telehealth, mHealth, telenursing, or digital self-care interventions for adults with T2DM involving nursing or nursing-relevant digital care component. Data were charted using the systematic table and synthesized through descriptive and thematic analysis.

Results

A total of 26 studies were included, representing diverse models such as app-based self-management, messaging platforms, telemonitoring, virtual education, EHR-supported discharge pathways, virtual group visits, and comprehensive telehealth. Nursing roles were categorized as digital education, clinical monitoring, care coordination, behavioral coaching, medication and follow-through support, peer-support moderation, and digital onboarding. Of the 26 included studies, nine (34.6%) were classified as nurse-led, five (19.2%) as nurse-supported, seven (26.9%) as multidisciplinary/team-based, and five (19.2%) had insufficient reporting to isolate a nurse-specific intervention role. Interventions featuring structured nurse involvement, individualized feedback, theory-informed behavior-change strategies, and escalation pathways demonstrated the most consistent improvements in self-care behaviors and selected glycemic outcomes, particularly HbA1c and fasting blood glucose.

Conclusion

Telehealth interventions for T2DM self-care should be regarded as hybrid care models rather than solely technology-based tools. The synthesized evidence supports nurse-mediated, theory-informed, and equity-sensitive digital diabetes care, while underscoring the need for improved reporting on nursing scope, digital engagement, clinical escalation, fidelity, cost, and implementation context.

Introduction

Type 2 diabetes mellitus (T2DM) remains a significant global public health challenge. The International Diabetes Federation estimates that 589 million adults aged 20 to 79 will have diabetes by 2025, with this number expected to rise to 853 million by 2050.¹ T2DM is a chronic, progressive disease that requires ongoing metabolic control, consistent behavioral management, and regular interaction with health services. Inadequate control increases the risk of microvascular complications such as retinopathy, nephropathy, and neuropathy, as well as macrovascular complications including cardiovascular disease, stroke, and peripheral arterial disease.^1,2 These complications reduce quality of life, increase disability, and place sustained pressure on patients, families, and healthcare systems. Continuous self-care is therefore central to T2DM management and complication prevention.^2,3

Diabetes self-care requires individuals to perform multiple and ongoing self-management activities, including healthy eating, physical activity, medication taking, glucose monitoring, risk reduction, and problem solving, while maintaining ongoing collaboration with the healthcare.⁴ Diabetes self-management education and support (DSMES) helps patients develop the knowledge, skills, confidence, and behavioral capacity needed to manage these activities.^2,5 However, many health systems still rely heavily on conventional face-to-face care. This model often depends on clinic attendance, scheduled appointments, and brief periodic consultations. Such an approach may be insufficient for a chronic disease that requires frequent feedback, ongoing behavioral reinforcement, and continuous support outside the clinical setting.^2,5

Conventional face-to-face diabetes care faces structural limitations, including geographic barriers, healthcare professional shortages, and fragmented support between clinical visits. To address the persistent gap between patient needs and health system capacity, digital health and telehealth have emerged as critical strategies, offering remote monitoring and virtual education.^6,7 Furthermore, recent advancements in digital therapeutics and artificial intelligence (AI)-enabled diabetes management—such as predictive glycemic algorithms and smart virtual assistants—are rapidly expanding this landscape by offering highly personalized, real-time patient support^8–10 These modalities transition diabetes management to a connectivity-based model, facilitating more frequent patient-provider interactions, individualized feedback, and adaptable support directly within patients’ homes and communities where many determinants of glycemic control arise.

Despite all the opportunities, evidence on telehealth interventions for T2DM self-care remains heterogeneous. Some studies report clinically meaningful reductions in glycated hemoglobin (HbA1c), fasting blood glucose, or other metabolic indicators.^6,11–13 Other studies show stronger effects on proximal outcomes, such as diabetes knowledge, self-efficacy, insulin adherence, foot-care behavior, dietary behavior, physical activity, or quality of life, without consistent improvement in glycemic outcomes.^14–17 Variations in outcomes can be attributed to differences in intervention type, delivery mode, duration, study design, baseline glycemic control, comparator conditions, patient characteristics, and the degree of integration with clinical services. App-based education, SMS reminders, telemonitoring, video consultation, social media support, virtual group visits, and hybrid nurse-led care represent distinct interventions, as their mechanisms, intensity, and implementation requirements vary considerably.^6,10–34

Nurses play a key role in digital diabetes care by transforming digital interactions into meaningfulclinical and behavioral support. In T2DM self-care interventions, they may serve as educators, care coordinators, behavioral coaches, remote monitors, digital facilitators, peer-support moderators, and clinical triage personnel.^{6,12,16,17,21,32–34} Within diabetes self-management education and support, nurses can assist patients in interpreting glucose data, strengthening self-management skills, identifying barriers to adherence, reinforcing behavior change, and facilitating coordination with the healthcare team when additional clinical management is required.^4,23 However, the extent of nursing involvement is not always reported with sufficient clarity. Several studies characterize interventions as nurse-led,^{6,10,16,23,25,27,29,32,33} while others describe nurses as supporters^{14,17,24,31,34} within broader multidisciplinary or technology-based programs. This distinction is significant because nurse-led care denotes greater autonomy, structured clinical responsibility, and potential decision-making authority. In contrast, nurse-supported care may focus on education, reminders, or technical assistance but typically lacks authority to escalate clinical care.^6,12,25,32

Theoretical clarity remains a significant concern in digital diabetes interventions. Theories such as Self-Efficacy Theory, Social Cognitive Theory, the Transtheoretical Model, Motivational Interviewing, and self-management frameworks provide potential explanations for how interventions affect confidence, motivation, problem-solving, adherence, and the maintenance of self-care behaviors.^35–37 When intervention design is not grounded in an explicit conceptual framework, it becomes challenging to elucidate the mechanisms through which, and the conditions under which, improvements in outcomes occur. Observed improvements may reflect the effects of individual intervention components, such as reminders, accountability, social support, nurse–patient contact, medication intensification, and self-monitoring, or may result from the synergistic effects of these components when delivered as part of a multifaceted service package.

Implementation and equity challenges contribute to the complexity of the global telehealth landscape for self-care of T2DM. The effectiveness of digital approaches relies on factors beyond the mere availability of devices or platforms. Key determinants include digital literacy, health literacy, internet connectivity, smartphone access, language, cultural context, privacy concerns, usability, patient age, disability, socioeconomic status, and the workload of healthcare professionals.³⁸ Rural populations, older adults, low-income groups, and individuals with limited digital skills are likely to benefit from digital care only when interventions are tailored to their specific circumstances. In contrast, technology-intensive interventions may exacerbate health inequities if they presuppose stable internet access, advanced digital literacy, or sustained engagement with digital applications. Therefore, digital diabetes care should be evaluated on the basis of feasibility, acceptability, reach, sustainability, and equity-sensitive implementation outcomes, rather than relying solely on clinical endpoints.^38,39

Previous systematic reviews and meta-analyses suggest that telehealth, mobile health, and other digital self-management interventions can improve glycemic control and support diabetes self-management among adults with T2DM, although their effects vary across intervention modalities, intensity, and design.^40–42 However, most have concentrated on clinical effectiveness, specific technologies, or general digital health outcomes. Limited attention has been paid to the integrated relationships among digital modality, nursing roles, conceptual frameworks, care models, and implementation contexts. This gap hinders a comprehensive understanding of how technology, nursing practice, behavioral mechanisms, and health-system readiness collectively support self-care in T2DM.

Accordingly, this scoping review sought to map the global landscape of telehealth interventions for T2DM self-care, with particular emphasis on nursing roles, care models, conceptual frameworks, and reported outcomes. The review addressed four primary questions: (1) which telehealth modalities have been employed to support T2DM self-care; (2) in what ways are nurses involved and how are their roles characterized; (3) which conceptual models inform these interventions; and (4) what clinical, behavioral, psychosocial, implementation, and equity-related outcomes have been documented. By integrating technology, nursing, theoretical frameworks, and outcomes, this review aims to clarify the current evidence base and identify priorities for future research, nursing practice, and health-system implementation in digital diabetes care.

Given the substantial heterogeneity in digital modalities, varying definitions of nursing roles, and diverse theoretical frameworks across the literature, a systematic review focused solely on clinical efficacy would be premature. A scoping review methodology is uniquely suited to map this complex, fragmented evidence base, clarify conceptual boundaries, and identify gaps in the literature. Accordingly, this scoping review sought to map the global landscape of telehealth interventions for T2DM self-care, with particular emphasis on nursing roles, care models, conceptual frameworks, and reported outcomes. The review addressed four primary questions: (1) which telehealth modalities have been employed to support T2DM self-care; (2) in what ways are nurses involved and how are their roles characterized; (3) which conceptual models inform these interventions; and (4) what clinical, behavioral, psychosocial, implementation, and equity-related outcomes have been documented.

Materials and Methods

Study Design

This scoping review aimed to systematically map the scope, characteristics, and knowledge gaps of nurse-led or nurse-facilitated telehealth interventions for self-care of T2DM. The review followed Arksey and O’Malley’s framework, the Joanna Briggs Institute (JBI) guidance, and the PRISMA-ScR reporting checklist.^43–46 A scoping design was appropriate because the evidence base was heterogeneous in intervention modality, nursing role, conceptual framework, study design, setting, and outcome measurement. In alignment with the iterative mapping nature of this methodology, the review protocol was not prospectively registered in registries such as the Open Science Framework (OSF) or PROSPERO.

Eligibility Criteria

Eligibility criteria were established using the Population-Concept-Context (PCC) framework. The population (P) included adults aged 18 years or older with T2DM. The concept (C) focused on telehealth, mHealth, telenursing, mobile applications, messaging platforms, virtual education, telemonitoring, or similar digital interventions involving nursing or nursing-relevant components within digital T2DM self-management care. The context (C) encompassed community, primary care, outpatient, home-based, transitional care, hospital discharge, rural, urban, and other global settings. Eligible empirical studies reported at least one clinical, behavioral, psychosocial, self-care, adherence, usability, feasibility, or implementation outcome.

Studies were considered eligible if they explicitly involved nurses, incorporated a multidisciplinary digital care model that included a nursing role, or evaluated telehealth interventions for T2DM self-management that were clinically relevant and could be implemented by nurses, even when the nursing role was not explicitly described in the article. Conversely, studies were excluded if they did not involve nurses or included care models or components that were not relevant to nursing practice, particularly in supporting self-management among patients with T2DM. When the intervention did not address T2DM self-care or diabetes self-management support, the publication was a review, editorial, commentary, conference abstract without full empirical data, protocol without results, non-peer-reviewed report, or the full text was unavailable. Mixed diabetes-type studies were included only when the intervention and outcome interpretation were relevant to T2DM or when T2DM participants comprised the majority.

Information Sources and Search Strategy

The search log documents searches conducted in PubMed/MEDLINE, EBSCOhost, Scopus, Taylor and Francis Online, ScienceDirect, and Sage Journals. The search strategy aimed to identify empirical studies examining telehealth interventions for T2DM self-care with nursing involvement. Additional studies were identified through manual screening of reference lists. The core search structure was: (type 2 diabetes or type 2 diabetes mellitus or t2dm) AND (nursing or nurse or Nursing-led Intervention OR Nursing intervention OR Nursing participation OR Nursing Involvement OR Nursing Roles) AND (Digital Health OR Digital Intervention OR Telehealth OR Telecare OR MHealth OR Telenursing OR Mobile Health) AND (Self-Care Management OR Self-care OR Self-Management).

Electronic databases were searched for English-language peer-reviewed articles published between January 1, 2021, to June 26, 2026, to capture the most recent advancements in digital health interventions for diabetes care. The final systematic search across all databases was completed on June 26, 2026.

Study Selection

All identified citations were collated and managed using Mendeley Reference Manager. Subsequently, the article screening and selection process was conducted using the Rayyan AI web application to facilitate blinded independent screening by the reviewers. Duplicate records were identified and removed prior to screening. Following deduplication, two reviewers independently screened the remaining records’ titles and abstracts to assess relevance to the review objectives and eligibility criteria. Studies that did not meet the inclusion criteria were excluded at this stage.

The full texts of potentially eligible studies were retrieved and assessed in detail by the same two reviewers. Each article was evaluated according to predetermined inclusion and exclusion criteria, including population, concept, context, study design, publication year, and relevance to the review question. Reasons for exclusion at the full-text stage were documented to maintain transparency in the selection process. Disagreements between reviewers were resolved through discussion, and if consensus was not achieved, a third reviewer made the final decision. The overall study selection process was reported using a PRISMA-ScR-style flow diagram, which detailed the number of records identified, duplicates removed, records screened, full-text articles assessed for eligibility, studies excluded with reasons, and final studies included in the review. This approach ensured that the study selection process was systematic, transparent, and reproducible.

Data Charting Process

A standardized data-charting matrix was developed to systematically extract study-level information from all included studies. Extracted variables included bibliographic information (author and year), country and setting, study design, sample size, participant characteristics and relevant clinical baseline characteristics, intervention characteristics, digital platform, modality and synchronicity, intervention duration and frequency, theoretical or conceptual framework, intervention provider, comparator, nursing involvement and specific nursing activities, outcomes, key findings, implementation considerations, and limitations relevant to generalizability.

The extracted data were subsequently organized at different levels of detail. Table 1 summarizes the core characteristics of the included studies, including study design, country and setting, sample size, participant characteristics, and relevant clinical baseline characteristics. Table 2 provides a regional synthesis of the digital intervention patterns observed across the included studies, including the predominant telehealth modalities, representative studies, and cautious contextual interpretations. Nursing involvement was further synthesized in Table 3, according to the reported intervention-delivery structure and professional responsibilities, distinguishing nurse-led, nurse-supported, multidisciplinary/team-based interventions, and studies in which a nurse-specific intervention role could not be clearly identified.

Table 1.

Characteristics of Included Studies

Study	Study Design	Country & Setting	Sample Size (N Text)	Participant Characteristics	Clinical Baseline
Tan et al 202211	Prospective single-arm cohort with age-stratified pre-post analysis	China Hospital-based diabetes follow-up supported by a doctor-nurse-patient Mobile Health Management System in China	464 enrolled 424 completed all scheduled 12-week follow-ups	Adults with type 2 diabetes Grouped as young 18–40 years, middle-aged 41–65 years, and elderly >65 years Mean age 55.0 ± 13.7 years.	Baseline FPG was high across groups: Young 178.38 ± 95.04 mg/dL, Middle-aged 180.00 ± 91.08 mg/dL, Elderly 174.24 ± 80.64 mg/dL.
Lee and Lee 202518	Quasi-experimental nonequivalent control group pre-post study	South Korea Community health promotion center, non-face-to-face diabetes self-management support	51 completed: Intervention n=27 Control n=24	Adults aged 30–70 years With type 2 diabetes, HbA1c ≥6.5%, Internet access through computer or mobile device; Mean age 58.4 ± 7.98 years; 74.5% women.	Baseline HbA1c was near 6.7% in both groups, Leaving limited room for HbA1c reduction; Diabetes duration averaged 5.44 ± 5.01 years
Tourkmani et al 202419	Prospective multicenter cohort pre-post intervention study	Saudi Arabia Three chronic illness centers affiliated with Prince Sultan Military Medical City, Riyadh	181 patients with uncontrolled type 2 diabetes	Adults with uncontrolled T2DM HbA1c ≥9%; 56.9% women; Mean age 58.64 ± 11.23 years; Diabetes duration 13.80 ± 8.55 years; 79.6% on insulin.	Mean baseline HbA1c 10.47 ± 1.23%; BMI 31.16 ± 6.43 kg/m2.
Park et al 202420	Parallel randomized pilot trial with qualitative interviews	South Korea Diabetes outpatient clinic at Pusan National University Yangsan Hospital	32 enrolled from 61 eligible Retention at 3 months 84.2% intervention 84.6% control	Adults aged 40–69 years with T2DM and HbA1c ≥7.0%, Using Android smartphones; Recruited from an outpatient clinic.	Participants had suboptimal HbA1c at eligibility; Groups were reported homogeneous at baseline.
Ory et al 202521	Three-arm randomized controlled comparative effectiveness trial	United States Rural and urban Texas communities; virtual DSMES and technology-based education support	189 randomized TBES n=68 vMMWD n=60 Combined n=61	Adults with T2DM, A1c ≥7.5%, Age ≥25 years, Smartphone/internet access, English-speaking; Median age 52 years; 77% female; 86% White; 39% Hispanic or Latino; 67 rural and 122 urban participants.	Baseline median A1c was 8.9%.
Roth et al 202522	Open-label cluster-randomized controlled multicenter pilot study	Austria Three Austrian outpatient diabetes educational group courses	15 participants analyzed Target sample was 73	Adults with type 2 diabetes attending outpatient educational group courses, Smartphone-capable, German-speaking Excluded type 1 diabetes Gestational diabetes Visual impairments.	Intervention group had higher diabetes self-management at baseline than control.
Pekmezaris et al 202523	Comparative effectiveness randomized controlled trial using a Zelen consent approach	United States Nine endocrinology or primary care outpatient practices and a safety-net hospital in a New York metropolitan area	240 adults with Hispanic or Latino background: DTM n=120, COM n=120	Underrepresented Hispanic or Latino adults with type 2 diabetes 65.4% women Mean age about 56.5 years 74% had high school education or less Many had low income.	Baseline A1C was 8.32% in both groups.
Crowley et al 20226	Active-comparator parallel-arm randomized clinical trial	United States Two Veterans Affairs health care systems	200 randomized: Comprehensive telehealth n=101, Telemonitoring plus care coordination n=99	Outpatients with persistently poorly controlled T2DM HbA1c ≥8.5% for at least 1 year despite clinic-based care Mean age 57.8 ± 8.2 22.5% women 72.0% Black 5.5% Hispanic or Latinx.	Baseline HbA1c 10.17% in both groups; mean BMI 34.8 kg/m2.
Shaban et al 202410	Quasi-experimental controlled pre-post study	Egypt Outpatient clinics at Cairo University Hospital	120 adults with type 2 diabetes: Intervention n=60, Control n=60	Adults with type 2 diabetes Aged >18 years Focus on older adults Intervention mean age 55.3 ± 3.2; Control mean age 57.1 ± 7.	Diabetes duration averaged about 7.4 years in intervention and 8.0 years in control.
White et al 202224	Randomized open-label single-center controlled trial stopped early	United States Hospital discharge pathway for insulin-requiring hospitalized patients with T2DM	158 enrolled: Discharge order set n=82 Enhanced standard care n=76 155 had discharge data	Hospitalized insulin-requiring patients with type 2 diabetes Mean age 51.7 ± 10.2 years 43% male 46.8% White Median diabetes duration 11 years; 81.9% Using insulin before admission.	Poorly controlled hospitalized diabetes population; BMI about 38.2 kg/m2.
Alsahli et al 202525	Two-phase cross-sectional and quasi-experimental pretest-posttest study	Egypt Kafr El Sheikh University Hospital	Phase I n=400 surveyed; Phase II n=100 selected for 4-week telenursing intervention	Patients with type I or type II diabetes; Mean age 49.7 ± 11.5 years; 33.8% diagnosed >10 years; 36.8% obese; 44% insulin only; High cardiac comorbidity.	Many had fasting high blood glucose symptoms; selected intervention group had low self-care practice knowledge.
Sungkapinyo et al 202514	Quasi-experimental two-group pretest-posttest study	Thailand Diabetes clinic in a tertiary hospital in southern Thailand	60 patients: Experimental n=30, Control n=30	Patients with uncontrolled type 2 diabetes Using insulin penfill for at least one year; HbA1c >7% Experimental mean age 54.90 ± 10.03, Control 58.23 ± 9.46; 53.3% Female in each group.	Baseline knowledgeinsulin adherence, and capillary blood glucose were not significantly different between groups.
Watcharanat 202515	Quasi-experimental non-equivalent control group study with three arms	Thailand Princess Maha Chakri Sirindhorn Medical Center and elderly diabetes management context	210 elderly individuals with diabetes: Mobile app n=70, Telephone follow-up n=70, Usual care n=70	Elderly patients with diabetes 61% female Group mean ages approximately 63.9 to 69.3 years.	No significant baseline differences in diabetes knowledge, health behaviors, or HbA1C among the three groups.
Esmailpour et al 202526	Randomized controlled trial	Iran Older adult type 2 diabetes care context in Iran	Abstract reports 102 randomized: Experimental n=51, Control n=51; Baseline table reports 45 per group analyzed	Older adults with type 2 diabetes Many had education at high school level or less Insufficient income, comorbidities, and oral medication use.	HbA1c was about 7.25% in experimental and 7.24% in control; FBS was 171 mg/dL and 181 mg/dL respectively.
Park and Park 202427	Randomized controlled trial	South Korea Korean public diabetes education center in the Seoul metropolitan area	105 older adults DiaNote n=53, Traditional logbook control n=52	Adults >65 years with type 2 diabetes, Smartphone ownership Intact cognition and communication Diabetes duration at least 1 year Excluded insulin users and severe medical conditions.	Mean age 72.70 years in DiaNote 74.12 years in control. Baseline HbA1c was 6.95 ± 0.87 in DiaNote and 6.83 ± 0.85 in control.
Romadlon et al 202528	Assessor-blinded three-arm randomized controlled trial	Indonesia Diabetes management center in East Java Province	84 recruited and randomized: DB-TEXT plus PSE n=28, DB-TEXT n=28, PEP n=28; 83 completed	Individuals aged 17–65 years with uncontrolled type 2 diabetes HbA1c >7% in previous three months Able to read Indonesian Owning a mobile phone More than 60% of participants were women.	Mean ages 54.5, 53.4, and 52.8 years across arms Diabetes duration about 6 years; baseline outcomes homogeneous.
Jafar et al 202329	Experimental controlled study with multilevel modeling	Indonesia Three primary care health centers	66 enrolled; 62 completed: Intervention n=29 Control n=33	People with type 2 diabetes from primary care Intervention group mostly rural Baseline groups differed in some residence distribution characteristics.	HbA1c was high, especially in the intervention group: intervention 9.56 ± 1.95%, control 8.66 ± 1.69%.
Dinh et al 202430	Single-arm pilot study	United States Five Midwestern community health centers during the COVID-19 pandemic	48 enrolled; 34 attended at least one virtual group visit; 21 completed 6-month survey	Adult patients with diabetes at community health centers A1C ≥8%, English or Spanish speaking, and at least two clinic visits.	Subgroup with baseline A1C ≥9% showed the clearest glycemic improvement.
Kobe et al 202212	Mixed-methods implementation study using RE-AIM; single-arm clinical implementation evaluation	United States Seven geographically diverse Veterans Health Administration sites serving rural patients	230 patients with clinic-refractory uncontrolled T2D	Rural patients with clinic-refractory uncontrolled T2D Average age 59 years 95% male 80% White 14% Hispanic or Latinx.	Model-estimated mean baseline HbA1c was 9.56%. Eligible population generally had HbA1c ≥8.5% despite VHA care.
Johnson et al 202331	52-week multisite parallel randomized controlled trial with longitudinal repeated measures	United States Web-based DSME/S study coordinated through Duke University and New York University; participants recruited from North Carolina and southern New England including New York	211 consented and randomized: LIVE n=102, Control website n=109	Adults with type 2 diabetes Age ≥21 years English literacy Home computer and broadband access, and telephone access Mean age 58.85 ± 10.1 years 59.2% female 60.2% White 93.8% non-Hispanic.	Mean HbA1c was 7.64% ± 1.79 Mean BMI was 33.51 ± 7.25. Many participants began near glycemic target Reducing room for HbA1c improvement.
Asante et al 202532	Single-blinded parallel-group two-arm randomized controlled trial	Ghana Diabetes Centre of Komfo Anokye Teaching Hospital, Kumasi, Ghana	98 randomized: Nurse-led mobile phone Intervention n=49, Standard care n=49	Adults with type 2 diabetes Baseline A1C >7% Mean age was 58.71 ± 11.04 in intervention 57.76 ± 10.96 in control Most were women Diabetes duration was about 10–11 years Most had basic education and low monthly income.	Baseline A1C was high and comparable: Intervention 10.14% ± 1.88 Control 10.28% ± 2.24. BMI was 27.71 ± 7.22 vs 27.45 ± 4.58.
Yaagoob et al 202416	Randomized controlled trial with baseline, 6-week, and 12-week follow-up assessments	Saudi Arabia Diabetic Education Clinic, Jizan Diabetes Center, Jazan, Saudi Arabia	80 randomized: intervention n=40, control n=40	Adults with type 2 diabetes aged 20–62 years; Mean age 40.55 ± 8.00 years 65% male 90% married All participants used WhatsApp 81.3% used oral hypoglycemic treatment.	Baseline HbA1c was high and balanced Intervention 10.31% ± 1.30 Control 10.41% ± 1.38. Baseline self-efficacy and knowledge were comparable.
Firdaus et al 202317	Single randomized two-group pretest-posttest controlled trial using per-protocol analysis	Malaysia Government-run primary clinic in Northern Malaysia	68 recruited; 58 included in final analysis after attrition and exclusions Final analysis included 29 per group.	Adults aged 18–65 years Uncontrolled diabetes Last two HbA1c readings >6.5% At least stage 1 risk for foot ulcer, Android phone ownership Ability to use a mobile app.	Baseline HbA1c was comparable: intervention 9.138 ± 1.956 and control 9.151 ± 2.237. Age, monthly income, and family history differed between groups and were controlled in analyses.
Kılınç İşleyen et al 202533	Parallel-group randomized controlled trial With post-test and 6-month follow-up	Türkiye Diabetes polyclinic of Bakırköy Dr. Sadi Konuk Education and Training Hospital, Türkiye	70 completed: Intervention n=36, Control n=34	Adults with type 2 diabetes Mean age 55.22 ± 7.07 in intervention 55.55 ± 8.31 in control Diabetes duration 13.58 ± 8.47 years in intervention 12.85 ± 9.64 years; in control Most participants used insulin plus oral therapy.	Baseline FBG was 217.46 ± 73.88 vs 215.00 ± 104.80. Baseline HbA1c was 9.53% ± 1.87 vs 9.31% ± 2.42. Groups were homogeneous on baseline characteristics and metabolic parameters.
Sjattar et al 202434	Quasi-experimental two-group interventional study With randomized group assignment Repeated measures	Indonesia Puskesmas Kendahe, Puskesmas Enemawira, and Puskesmas Kuma in Sangihe Islands Regency, Indonesia	84 participants: Intervention n=42, Control n=42	People diagnosed with diabetes mellitus Aged ≥40 years Able to read and write Cooperative, and owning a cell phone. The intervention and control groups were homogeneous on most baseline characteristics and FBG.	Baseline FBG was 210.88 ± 50.54 mg/dL in intervention 206.36 ± 51.99 mg/dL in control. Baseline self-care knowledge was 17.95 ± 4.54 intervention 18.83 ± 3.97 control
Zheng et al 202613	12-week nonblinded multicenter randomized controlled trial 2-year FBG follow-up	China Three community health service centers in Chongqing, China	600 randomized: WeChat intervention n=300 Control n=300 556 completed 12-week follow-up	Community-dwelling adults with type 2 diabetes Receiving contracted family physician services, Able to use WeChat, Face-to-face interview, and smartphone functions. Follow-up sample was n=278 per group; Mean age about 59 years; 57% women.	Baseline HbA1c was 7.39% ± 1.53 in intervention and 7.44% ± 1.46 in control. Baseline FBG was 8.12 ± 2.57 mmol/L intervention 8.39 ± 2.73 mmol/L.caontrol

Table 2.

Regional Distribution of Telehealth Patterns in Included Studies

Region	Observed Telehealth Pattern in Included Studies	Representative Studies	Cautious Interpretation
North America	Telemonitoring, EHR/order-set support, virtual group visits, immersive virtual education	Pekmezaris et al;23 Crowley et al;6 White et al;24 Dinh et al;30 Kobe et al;12 Johnson et a l(31)	Included studies from this region more often used health-system integrated telehealth infrastructure
East Asia	App-based self-care, WeChat, non-face-to-face education, older-adult digital self-care	Tan et al;11 Lee and Lee;18 Park et al;20 Park and Park;27 Zheng et al13	Several studies used platform-based education, self-monitoring, and community health support
Southeast Asia	LINE, WhatsApp, mobile apps, app-based coaching, low-bandwidth telenursing	Sungkapinyo et al;14 Watcharanat;15 Jafar et al;29 Firdaus et al;17 Sjattar et al34	Included studies reflected mobile access, local language needs, and variable internet conditions
Middle East and North Africa	WhatsApp, hybrid telemedicine, digital nursing education, telenursing follow-up	Tourkmani et al;19 Shaban et al;10 Alsahli et al;25 Yaagoob et a l.16	Remote nursing support was used to address continuity and self-care education needs
Sub-Saharan Africa	Nurse-led mobile phone intervention	Asante et al;32	Low-bandwidth mobile phone care appeared appropriate in the included Ghanaian setting

Table 3.

Classification and Functional Synthesis of Nursing Involvement in Telehealth Interventions for T2DM Self-Care

Nursing Involvement	Operational Definition	n (%)	Representative Studies	Dominant Nursing Roles	Common digital modalities
Nurse-led	Nurses were explicitly identified as the primary interventionists or delivered/coordinated the core patient-facing components of the digital intervention.	9 (34.6)	Crowley et al;6 Yaagoob et al;16 Pekmezaris et al;23 Shaban et al;10 Alsahli et al;25 Park & Park;27 Jafar et al;29 Asante et al;32 Kılınç İşleyen et al33	DSMES/digital education; behavioral coaching; motivational interviewing; telemonitoring; individualized feedback; self-management support; RN case management; follow-up and care escalation where specified	Telephone/mobile calls; telemonitoring; WhatsApp; digital nursing platforms; mobile apps plus RN calls
Nurse-supported	Nurses had an identifiable patient-facing or intervention-support role, but the core intervention was substantially technology-, researcher-, or system-driven, or primary nursing leadership could not be established.	5 (19.2)	Sungkapinyo et al;14 Firdaus et al;17 White et al;24 Johnson et al;31 Sjattar et al34	Education reinforcement; adherence support; app/digital facilitation; remote follow-up; monitoring; telephone support	EHR + telephone; LINE; mobile apps; WhatsApp; cellular telenursing/virtual environment
Multidisciplinary/ team-based	The intervention was intentionally delivered by two or more professional disciplines or an interdisciplinary clinical team, such that overall delivery could not reasonably be attributed primarily to nursing.	7 (26.9)	Tan et al;11 Kobe et al;12 Zheng et al;13 Tourkmani et al;19 Ory et al;21 Roth et al;22 Dinh et al30	Shared education; telemonitoring; patient-data review; care coordination; counseling; referral/escalation; collaborative DSMES	Doctor–nurse–patient mHealth; integrated telehealth; group training; virtual group visits; WeChat/community digital platform
Nurse role not specifically identifiable	The intervention was reported within a nursing-related or digital self-management context, but the professional identity of the intervention deliverer or urse-specific patient-facing contribution could not be clearly established from the published report.	5 (19.2)	Watcharanat;15 Lee & Lee;18 Park et al;20 Esmailpour et al;26 Romadlon et al28	Intervention development/evaluation; digital education or support where professional attribution was insufficiently reported	Mobile apps; virtual education; SMS/text messaging; LINE/SNS-supported self-management

Detailed intervention-level data are provided in Supplementary Table S1, including the theoretical or conceptual framework, digital intervention characteristics, platform, modality taxonomy and synchronicity, intervention duration and frequency, provider, comparator, reported nursing involvement, and specific nursing activities. This layered presentation was used to retain detailed study-level information in the supplementary material while presenting the main patterns and cross-study synthesis in the main manuscript. Data extraction was conducted by one reviewer and verified by at least one additional reviewer for accuracy and consistency.

Data Items and Narrative Synthesis

Data synthesis employed both descriptive mapping and narrative synthesis. Descriptive mapping provided an overview of countries, study designs, sample sizes, intervention modalities, platforms, theoretical frameworks, nurse roles, and outcome domains. Narrative synthesis organized findings according to telehealth modality, regional patterns, nursing roles, theory-to-mechanism alignment, clinical outcomes, behavioral outcomes, psychosocial outcomes, implementation barriers, equity considerations, and methodological limitations. The analysis was qualitative and interpretive. A meta-analysis was not performed due to substantial heterogeneity across study design, intervention dose, comparator conditions, follow-up duration, and outcome measurement.

Formal critical appraisal was not applied as an exclusion criterion, in alignment with established scoping review methodology. Methodological limitations were documented narratively to facilitate cautious interpretation of outcome patterns. These limitations encompassed sample size, follow-up duration, randomization, comparator strength, attrition, reliance on self-reported outcomes, fidelity reporting, and clarity regarding nursing role autonomy.

Results

Study Selection

A total of 1907 records were identified across six electronic databases. Following the removal of 87 duplicates, 1820 records remained for title and abstract screening. Of these, 1731 records were excluded because they were reviews, lacked relevance to the topic, focused on other conditions, or did not match the specified population or intervention concept. 89 full-text articles were assessed for eligibility, resulting in the exclusion of 68 articles. This process yielded 21 studies from database searches. Manual searching identified five additional eligible studies. Consequently, the final review included 26 empirical studies (Figure 1).

Figure 1.

PRISMA flowchart.

Study Characteristics and Scope of the Evidence

The final extraction matrix included 26 empirical studies examining telehealth, digital nursing, telemonitoring, virtual education, mobile applications, messaging platforms, and other technology-enabled care models for adults with T2DM (Table 1). The combined sample comprised approximately 3842 participants. The evidence base was geographically diverse, encompassing studies from China, South Korea, Saudi Arabia, the United States, Austria, Egypt, Thailand, Iran, Indonesia, Ghana, Malaysia, and Turkiye. Methodological heterogeneity was observed, with 15 randomized trials, 7 quasi-experimental or non-randomized intervention studies, 2 prospective cohort or pre-post studies, 1 single-arm pilot study, and 1 mixed-methods implementation study. Supplementary Table S1 provides detailed study characteristics.

The included studies investigated a range of telehealth care models, such as doctor-nurse-patient mobile health management, remote diabetes self-management education, hybrid in-person and telemedicine care, automated personalized app-based self-care, comparative digital diabetes self-management education and support (DSMES) strategies, app-supported diabetes group training, culturally adapted telemonitoring, comprehensive telehealth, digital nursing education, electronic discharge order sets with nurse follow-up calls, telenursing education, LINE-based insulin support, and mobile application-based self-management.^{6,10,11,14,18–26} Further evidence included mobile health self-care training, older-adult digital self-care, personalized text messaging with peer support, app-based coaching, virtual diabetes group visits, rural intensive telehealth implementation, immersive virtual learning environments, nurse-led mobile phone coaching, WhatsApp-based DSMES, diabetic foot-care mobile support, tele-nursing motivational interviewing, low-bandwidth mobile telenursing, and WeChat-based community health education.^{12,13,16,17,26–34} Intervention details, theoretical frameworks, and nurse-role coding are shown in Supplementary Table S2.

Across the matrix, active-comparator and multicenter randomized trials represented the strongest study designs. Several studies offered feasibility, implementation, or contextual insights rather than definitive causal evidence. This distinction is important, as findings from single-arm, cohort, or quasi-experimental studies require cautious interpretation unless supported by concurrent control groups, adequate randomization, sufficient follow-up, and robust fidelity reporting.

Telehealth Modality Profile and Regional Patterns

The distribution of technologies revealed four primary telehealth modality groups: synchronous or hybrid telehealth, mobile application-supported self-management, messaging or social-media-supported self-care, and telemonitoring or electronic health record (EHR)-enabled clinical workflow models (Table 2). These categories were not mutually exclusive, as many studies integrated app-based education, tele-follow-up, phone calls, SMS, social media, and clinical monitoring within a single intervention. The subsequent regional patterns reflect only the included studies and should not be considered a definitive global comparison. In North American studies, telehealth models more frequently incorporated infrastructure-intensive designs, such as telemonitoring, virtual group visits, EHR-supported discharge planning, comprehensive telehealth, and immersive virtual environments.^{6,23,24,30,31} Within East Asian and Southeast Asian studies, several interventions used familiar consumer-facing tools, such as mobile applications, WeChat, LINE, WhatsApp, SMS, and app-based self-recording systems,^{11,13–15,17,18,20,27–29,34} In resource-limited or geographically dispersed settings, low-bandwidth phone or messaging-based telehealth was contextually relevant because it reduced dependence on broadband infrastructure.^32,34

The most robust telehealth studies extended beyond replicating clinic appointments remotely. These interventions integrated sustained human contact, patient-generated data, structured feedback, and nurse-mediated interpretation. Crowley et al⁶ reported greater reductions in HbA1c with comprehensive telehealth than with telemonitoring combined with care coordination, suggesting that clinical actionability and intervention intensity may be critical factors. Kobe et al¹² showed that an intensive telehealth model could be implemented in rural Veterans Health Administration sites, with sustained within-cohort HbA1c improvement, although the design was implementation-oriented rather than randomized.

Nurse Roles and Involvement

Nursing involvement varied substantially across the digital interventions for T2DM self-care. Across studies with identifiable nursing activities, reported functions included digital diabetes education, clinical monitoring, behavioral coaching, digital facilitation, care coordination, medication or adherence support, peer-support facilitation, and digital onboarding. Classification as nurse-led indicates primary responsibility for delivering or coordinating the core patient-facing intervention components and does not necessarily imply independent prescribing authority or autonomous clinical decision-making. Detailed intervention-level nursing activities are presented in Supplementary Table S1, while their functional synthesis is summarized in Table 3.

Based on the reported intervention-delivery structure and professional responsibilities, the 26 included studies were classified into four categories. Nine studies (34.6%) were classified as nurse-led, in which nurses were explicitly identified as the primary interventionists or delivered or coordinated the core patient-facing components of the digital intervention. Five studies (19.2%) were classified as nurse-supported, in which nurses had identifiable roles in education, counseling, monitoring, digital facilitation, or follow-up but were not clearly the primary providers of the overall intervention. Seven studies (26.9%) used multidisciplinary or team-based models in which nursing activities were integrated with those of other healthcare professionals. In five studies (19.2%), a nurse-specific intervention role could not be clearly identified from the published report.

Nurse-led interventions demonstrated several forms of direct patient-facing engagement. For example, Crowley et al⁵ used nurses to deliver multiple components of comprehensive telehealth, including telemonitoring, self-management support, diet and activity support, and medication-related support. Pekmezaris et al²³ incorporated culturally congruent nursing into diabetes telemonitoring, while Asante et al³³ used trained nurses to deliver a staged mobile-phone self-management intervention. Yaagoob et al¹⁶ evaluated WhatsApp-based DSMES delivered by an expert diabetes health-education nurse, and Kılınç İşleyen et al³⁴ used nurse-delivered telephone motivational interviewing. These interventions illustrate nursing functions extending beyond technical facilitation to structured education, behavioral support, monitoring, individualized feedback, and longitudinal follow-up.

In nurse-supported models, nursing activities complemented technology-, researcher-, or system-mediated components. These roles included reinforcement of diabetes education, remote follow-up, adherence support, digital facilitation, and assistance with self-management monitoring. For example, Firdaus et al¹⁷ combined a mobile application with face-to-face and WhatsApp-supported components involving a diabetes nurse educator, whereas White et al²⁵ integrated postdischarge nursing telephone support with an electronic discharge workflow. Because intervention delivery in these studies was distributed across digital and human components, the observed outcomes should not be attributed exclusively to nursing.

Multidisciplinary or team-based models embedded nursing within broader interprofessional digital-care pathways. Tan et al,¹¹ for example, used a doctor–nurse–patient mobile health management system in which nurses provided education, follow-up, data collection, and coordination, while endocrinologists provided medication guidance. Similarly, other team-based interventions combined nursing-related activities with those of physicians, dietitians, diabetes educators, pharmacists, coaches, or other clinical staff. In these models, the independent contribution of nursing generally could not be separated from the overall intervention.

Finally, five studies were retained in a distinct category because nurse-specific intervention delivery could not be clearly identified from the published reports. This category was applied when nursing-related context or involvement was present but the professional identity of the intervention deliverer or the patient-facing contribution specifically attributable to nurses was insufficiently reported. Maintaining this distinction avoids attributing intervention delivery or outcomes to nursing solely on the basis of author affiliation, research setting, or disciplinary context.

Theory-Informed Narrative Synthesis

The theoretical grounding of the interventions was uneven. Approximately one-third of the matrix records were explicitly theory-based, another subset was partially theory-informed or organized around a clinical practice framework, and a substantial group had no explicit behavioral theory. This is important because theoretical opacity limits interpretation of why an intervention changed behavior, adherence, or glycemic outcomes. Detailed coding is provided in Table 4.

Table 4.

Theory-To-Outcome Alignment in Included Studies

Theoretical Alignment	Representative Studies	Dominant Outcomes	Narrative Interpretation
Self-efficacy or Social Cognitive Theory	Lee and Lee;18 Yaagoob et a16 l; Johnson et a l(31); Zheng et al13	Self-efficacy, self-management, social support, selected glycemic outcomes	Relevant for explaining behavioral pathways, but clinical superiority depends on comparator and intensity
Transtheoretical Model	Asante et al;32 elements in Firdaus et al17	A1C, self-management, foot care, dietary behavior	Useful for staged behavior change and nurse follow-up
Motivational Interviewing	Kilinc Isleyen et al33	Self-efficacy, self-management, FBG, triglycerides, short-term HbA1c	Compatible with nursing communication and autonomy support
IMB, ADCES7, or self-care frameworks	Park and Park;27 Romadlon et al;28 Firdaus et al17	Self-care skills, quality of life, adherence, diet and activity behavior	Helpful for organizing DSMES components and self-care targets
No explicit behavioral theory	Crowley et al;6 Kobe et al;12 White et al;24 Tourkmani et al19	Clinical or implementation outcomes	Clinical effects may occur, but causal mechanisms remain harder to isolate

Self-efficacy theory and Social Cognitive Theory were visible behavioral mechanisms in several interventions. Lee and Lee¹⁸ combined self-efficacy theory with health literacy in a non-face-to-face program and observed improvements in fasting blood glucose, self-efficacy, social support, and self-management behavior. Yaagoob et al¹⁶ used Bandura’s self-efficacy theory in WhatsApp-supported DSMES and reported improvements in self-efficacy, self-management, and education. Another similar study by Zheng et al¹³ Used social cognitive theory combined with Health belief models in WeChat-based intervention which showed superior HbA1c and fasting blood glucose outcomes and sustained lower fasting blood glucose at 2-year follow-up Johnson et al³¹ used Social Cognitive Theory and social support concepts in an immersive virtual environment, but did not show superior clinical outcomes compared with an active web-based control.

Stage-change and motivational frameworks proved especially compatible with nurse-led telecoaching. Asante et al,^32,36 used the transtheoretical model to structure an initial workshop and longitudinal nurse calls, aligning intervention components with stages from awareness-raising to maintenance. Kilinc Isleyen et al³³ used tele-nursing motivational interviewing and reported improvements in self-efficacy, self-management, fasting blood glucose, triglycerides, and short-term HbA1c. These outcomes support a careful interpretation that theory-informed nurse communication may be behaviorally important, although superiority cannot be inferred across all designs.

Some clinically promising interventions did not clearly report a classic behavioral theory but were embedded in strong clinical systems. Crowley et al⁶ and Kobe et al¹² embedded telehealth within clinical staffing, telemonitoring, medication support, and implementation infrastructure. However, the lack of theoretical specification makes it difficult to determine whether outcomes were driven by reminders, monitoring, medication intensification, accountability, nurse contact, or the combined service package.

Outcome Patterns

Category I: Clinical Outcomes

Clinical outcomes were most frequently reported as HbA1c, fasting blood glucose, fasting plasma glucose, or capillary blood glucose. The clearest HbA1c signals came from comprehensive or integrated models. Crowley et al⁶ reported greater HbA1c reduction with comprehensive telehealth than with telemonitoring plus care coordination Kobe et al¹² reported sustained within-cohort HbA1c reductions in a rural implementation cohort, although causal inference was limited by the absence of a randomized comparator. Zheng et al¹³ showed significant glycemic benefit in a large multicenter WeChat-based trial. Tourkmani et al¹⁹ reported a large pre-post reduction in HbA1c after hybrid in-person and telemedicine care, but the single-cohort pre-post design limits causal interpretation.

Category II: Behavioral and Adherence Outcomes

Behavioral, along with adherence outcomes, included self-management behaviors, medication adherence, insulin adherence, foot care, dietary behaviors, physical activity, glucose monitoring, and self-care skills. These outcomes were more consistently positive than clinical outcomes. LINE-based insulin support improved insulin knowledge and observance behavior but did not produce a significant between-group difference in capillary blood glucose.¹⁴ The Diabetic Care App improved foot care and dietary behavior, but not fasting blood glucose.¹⁷ WhatsApp-based DSMES improved self-efficacy, education, and self-management.¹⁶ This pattern suggests that digitaltreatments may first shift proximal behavioral mechanisms, while clinical outcomes may require longer duration, higher baseline risk, medication adjustment, or stronger clinical integration.

Category III: Psychosocial and Older-Adult Outcomes

Psychosocial outcomes included self-efficacy, social support, depression, distress, quality of life, and fatigue. Older-adult interventions showed promise but also exposed a design gap. Watcharanat¹⁵ found that a mobile application group outperformed telephone follow-up and standard care in knowledge, health behaviors, and HbA1c among older adults. Park and Park²⁷ showed that DiaNote plus phone visits benefited blood sugar control and were comparable to traditional logbooks for some self-care outcomes. These conclusions support older-adult telehealth but also indicate the need for deliberate usability adaptation, training, and nurse-assisted onboarding.

Category IV: Implementation Outcomes, Frameworks, Facilitators, and Barriers

Implementation outcomes revealed a complex landscape of barriers and facilitators. Patient-level barriers and implementation considerations included digital literacy and familiarity, health-literacy challenges, technology acceptance, and requirements for access to or familiarity with digital devices and platforms,^{15,16,27,29–32} provider-level barriers (eg, increased workload, staffing limitations, and training needs),^12,30 technology-level barriers (eg, limited access to suitable digital devices, reliable internet connectivity, and technical support),^30,31 organizational-level barriers (eg, staffing constraints, workflow demands, and dependence on existing digital-health infrastructure)^12,30 Conversely, facilitators for successful digital self-care delivery included structured nurse-assisted onboarding, active family or caregiver involvement, culturally tailored content, and user-centered intuitive designs.

Despite these insights, the explicit use of formal implementation frameworks to systematically evaluate these factors remains limited. For instance, only a few studies, such as Kobe et al,¹² utilized established models like the RE-AIM framework to assess reach, effectiveness, adoption, implementation, and maintenance. Expanding the use of structured implementation frameworks—such as RE-AIM or the Consolidated Framework for Implementation Research (CFIR)—is essential to systematically map these facilitators and barriers across diverse health systems.

Furthermore, equity signals were particularly important in Ghana, Indonesia, China, Saudi Arabia, and studies involving Latino or older-adult populations. Asante et al³² addressed low-resource diabetes self-management through mobile phone delivery in Ghana. Sjattar et al³⁴ used ordinary cell phone-based telenursing because internet-dependent interventions were unsuitable for some island communities in Indonesia. Pekmezaris et al²³ addressed cultural and linguistic adaptation for underrepresented Latino adults in the United States. These examples indicate that digital health equity is a core determinant of intervention feasibility and effectiveness, not a secondary concern.

The methodological profile was mixed. Only a small proportion of studies could be considered stronger evidence of clinical effectiveness. Many studies had small samples, short follow-up, self-reported outcomes, non-randomized designs, lack of blinding, attrition, feasibility-related early stopping, mixed diabetes-type populations, or absence of active comparators. This is consistent with scoping review guidance: such reviews can map the breadth of evidence but should avoid definitive effect claims when methodological heterogeneity and bias signals are substantial.^45,46 The outcome matrix and barriers coding are shown in Supplementary Tables S2 and S3.

Discussion

Principal Findings

This scoping review demonstrates that telehealth interventions for T2DM now extend beyond basic online education. Current approaches include hybrid clinical services, app-based self-management, messaging and social media support, telemonitoring, EHR-supported discharge planning, virtual environments, and nurse-led behavioral coaching.^6,10–34 The primary finding is that technology alone does not account for the most significant outcome patterns. The most compelling studies integrated digital tools into structured nursing practices, theory-driven behavior change strategies, individualized feedback, and clinical workflows that could respond to patient-generated data.^6,12,13,32 The evidence indicates a hierarchical framework of telehealth maturity. At the most basic level, interventions provide education or reminders. Intermediate-level interventions facilitate self-recording, social support, adherence prompts, and feedback. Advanced interventions incorporate patient-generated data with nurse interpretation, care coordination, medication support, and escalation pathways. Higher-intensity models were more likely to improve HbA1c, especially among participants with poor baseline glycemic control. However, this conclusion should be interpreted with caution due to variations in study designs and comparators.^6,12,13,19

Integrating Nursing Roles within Hybrid Digital Care

This scoping review indicates that effective digital diabetes care rarely depends on a single platform. Rather, studies commonly integrated phone calls, web portals, mobile applications, messaging platforms, telemonitoring, and virtual visits within coordinated care models.^6–31 The key issue was therefore not the technology itself, but whether digital contact was linked to continuity, feedback, accountability, and clinical follow-up. This finding supports viewing mobile and messaging platforms as telehealth channels when they facilitate diabetes education, monitoring, communication, and follow-up. However, their clinical value depends on integration with a structured care process rather than standalone use.

Within this architecture, nursing roles were central to converting digital contact into clinical action. Across studies, nurses acted as clinical monitors, behavioral coaches, care coordinators, medication-adherence supporters, digital facilitators, and peer-support moderators.^{12,13,17,18,25,26,33–35} The strongest interventions combined education with interpretation of digital data, motivational communication, and structured follow-up. Yet nurse autonomy varied substantially: some studies described nurses with responsibilities for triage, escalation, medication adjustment under protocol, documentation of patient-generated data, and multidisciplinary coordination, whereas others limited nurses to education and reminders. This variation reduces reproducibility and obscures which nursing functions drive outcomes. Future studies should therefore specify the scope of nursing practice and decision-making authority within the digital care pathway. To synthesize these mechanisms, a conceptual visual summary illustrating how nursing roles integrate with telehealth modalities and theoretical frameworks to influence outcomes is presented in Figure 2.

Figure 2.

Conceptual synthesis of telehealth modalities, reported nursing involvement, theories and behavior-change approaches, and outcome domains in T2DM self-care. Implementation and equity factors shape intervention delivery. The arrangement represents conceptual relationships rather than tested causal pathways; nursing-role attribution and outcome findings varied across studies.

Theoretical Clarity Strengthens Behavioral Interpretation

Explicit behavioral theory was not universal, even though digital T2DM care is primarily a behavior-change intervention. Self-efficacy theory, Social Cognitive Theory, the transtheoretical model, motivational interviewing, and self-management frameworks helped explain how interventions may influence adherence, self-care, confidence, and maintenance.^{16–18,31–33,35–37} These models matched closely with nurse-led communication because they focused on patient agency, skill development, problem-solving, and sustained behavior. Atheoretical interventions may still produce clinical benefit when embedded within robust clinical systems, but they create uncertainty about mechanisms of change.^6,12

Equity and Implementation Readiness Should Be Core Outcomes

The review also shows that digital diabetes care is highly context-sensitive. Interventions requiring smartphones, broadband internet, or continuous app engagement may be difficult for rural, island, older, low-income, or low-literacy populations. Conversely, low-bandwidth phone calls or familiar messaging platforms may be more feasible in resource-limited contexts.^32,33 This supports current diabetes care guidance emphasizing DSMES, technology, personalization, and attention to social determinants of health.^2,3 The field should move from technology-first design to equity-sensitive design. This means selecting the digital channel based on patient capability and local infrastructure, not solely on innovation appeal. It also means measuring digital competence, usability, engagement burden, privacy risk, cost, and family or caregiver support as part of intervention evaluation.

Implications for Practice

Implications for Nursing Practice and Informatics

Clinical services must explicitly define digital nursing roles. At a minimum, digitally enabled type 2 diabetes mellitus (T2DM) care requires nurses to educate patients, facilitate technology onboarding, assess self-management barriers, review patient-generated data, document clinical findings, identify clinical deterioration, and appropriately escalate concerns. Advanced programs should prepare nurses to serve as telehealth care coordinators and clinical decision-support professionals, rather than as technical operators. Training should encompass remote glucose monitoring, risk triage, motivational interviewing, documentation of digital encounters, privacy and data governance for messaging platforms, interpretation of patient-generated health data, and culturally adaptive digital communication. Mastery of these competencies is essential for the safe scaling of digital diabetes care.

Implications for Clinical Guidelines and Service Design

Clinical guidelines and service protocols should clearly differentiate between app-based education, messaging support, telemonitoring, telemedicine visits, and comprehensive nurse-mediated telehealth. These modalities are not equivalent and should be treated distinctly. Interventions involving clinical monitoring and nurse-mediated action pathways may be more appropriate for individuals with poorly controlled or high-risk T2DM, while app-based or messaging interventions may better support behavioral reinforcement and self-management education. Service design must include escalation protocols, documentation standards, privacy safeguards, and clearly defined boundaries for nurse-led medication or clinical support. Social media and messaging platforms, including WhatsApp, LINE, Telegram, and WeChat, should be regulated by explicit rules regarding response times, confidentiality, emergency escalation, and moderation.

Implications for Research

Future RCTs should compare technology-only or message-only digital care, nurse-led hybrid telehealth, and usual or enhanced standard care. Such designs will help isolate the added value of nursing involvement beyond technology and clarify which intervention components drive observed changes. Future trials should report digital engagement metrics, nurse contact frequency, intervention fidelity, adverse events, usability, costs, access to digital technology, and equity moderators. Key outcomes should include HbA1c, fasting blood glucose, medication adherence, self-care behaviors, diabetes distress, self-efficacy, quality of life, retention, and implementation feasibility. Future studies should be explicit about theoretical frameworks and mechanisms of action. Investigators must specify whether interventions target self-efficacy, health beliefs, motivation, behavioral skills, social support, patient activation, or medication intensification, and test whether these mechanisms mediate clinical outcomes.

Limitations of This Study

Several limitations should be acknowledged. First, the final search date and complete database-specific strategies require confirmation against the original search log prior to submission. Second, heterogeneity in study design, comparator strength, follow-up duration, digital modality, and outcome measurement precluded pooled estimation of effectiveness. Third, reporting of nurse involvement was inconsistent, with many studies failing to specify whether nurses provided education, remote monitoring, protocol-based triage, medication support, escalation, or autonomous clinical decision support. Fourth, reliance on self-reported self-care outcomes or short-term metabolic indicators was common. Fifth, publication and language restrictions may have excluded relevant non-English or unpublished implementation evidence from low-resource settings. Consequently, this review should be interpreted as a map of evidence, mechanisms, and implementation gaps rather than a definitive review of effectiveness. Future reviews may employ systematic review or meta-analytic designs when the evidence base becomes more methodologically comparable.

Conclusion

This scoping review demonstrates that telehealth interventions for T2DM self-care are globally diverse and increasingly hybrid. The evidence base encompasses synchronous telehealth, mobile applications, messaging and social media platforms, telemonitoring, electronic health record (EHR)-supported discharge tools, virtual group visits, and immersive virtual environments. Across these models, the primary pattern is not the technology itself but rather the clinical and behavioral frameworks that support its implementation.

Digital approaches are most effective when technology is integrated within structured nursing roles, theory-informed behavior change strategies, individualized feedback mechanisms, and a health system prepared for digital integration. While app-only and messaging-only interventions can enhance reach, knowledge, and self-care behaviors, nurse-mediated telehealth is better positioned to translate digital contact into clinical monitoring, behavioral support, care coordination, and actionable diabetes management.

In conclusion, this scoping review highlights that nurse-mediated telehealth interventions significantly enhance T2DM self-care and glycemic outcomes when integrated with behavioral theories. However, the successful implementation of these digital care models is frequently impeded by significant barriers, including limited digital literacy, restricted internet access, technology anxiety, and cultural constraints among certain patient demographics. To overcome these challenges, future research should progress beyond descriptive evaluations and pilot studies by prioritizing high-quality randomized controlled trials (RCTs). These rigorous trials are essential to test which combinations of nursing roles, telehealth modalities, behavioral theories, clinical workflows, and equity-sensitive designs yield sustained improvements. Furthermore, subsequent studies must critically address the cost-effectiveness and long-term sustainability of these digital interventions to support broader health system adoption. In practice, these findings support a transition toward nurse-led hybrid digital diabetes care, underpinned by formal nursing informatics competencies, clearly defined escalation protocols, and patient-centered technology design.

Data Sharing Statement

No new primary data were generated for this review. The extraction matrix used for synthesis is available in the supplementary tables included with this manuscript. The complete database-specific search strategies and PRISMA-ScR checklist should be submitted as supplementary files.

Ethics Approval and Consent to Participate

This is not applicable. This scoping review used data from previously published studies and did not involve recruitment of human participants.

Disclosure

All authors declare that they have no conflicts of interest in this work.