Digital health for rural diabetes care: Implementation experience from China and India

Table 1 Comparing context for digital diabetes care in China and India, 2021-2023

Domain	China	India	Ref.
Population, 2023			World Bank Development Indicators[45]
Total population (millions)	1410	1438
Population aged > 65 years (%)	14.2	6.8
GDP per capita (current United States $)	12614	2480
Gini coefficient	46.5	32.8
Diabetes prevalence, 2022			NCD Risk Factor Collaboration[1]
Adults with diabetes (millions)	148	212
Percentage of global diabetes cases (%)	18	26
Age-adjusted prevalence (%)	14.3	13.9
Rural context, 2021			World Bank Development Indicators[45]
Rural population (millions)	499	915
Rural population (% of total)	35.4	63.6
Rural population growth (annual%)	-2.9	0.11
Digital infrastructure, 2023
Rate of internet users, national estimates (%)	77.5	51.5	Statista[46,47]
Rate of internet users, rural areas estimates (%)	61.8	53	Cyberspace Administration of China[48], World Bank[49]
Healthcare System Context, 2021			WHO Global Health Observatory[44]
Life expectancy at birth	77.1	70.2
Physician density (per 10000 population)	25.2	7.3
Nurse density (per 10000 population)	33	17.3
Out-of-pocket expenditure (% of CHE)	34	50
Government health expenditure (% of GDP)	2.91	1.12
Premature mortality from NCDs (% probability)	16	22

NCDs: Non-communicable diseases; GDP: Gross domestic product; CHE: Current health expenditure.

Table 2 Comparison of key digital health intervention studies for diabetes in rural China and India

Study	Sample size	Intervention description	Primary outcome(s)	Key results
Interventions in China and India
SimCard[9]	2086	Smartphone-based decision support for community health workers	Antihypertensive medication use	Medication use increased by 24.4% in China and 26.6% in India
Interventions in China
ROADMAP[8]	19601	mHealth-enabled hierarchical diabetes management	HbA1c control	Mean HbA1c difference: -0.30%; Better effect in rural areas
SMARTDiabetes[12]	2072	Self-management app with family health promoter support	Proportion achieving multiple targets (HbA1c, BP, LDL)	Mean HbA1c difference: -0.33%; Effective in rural areas but not in urban settings
Interventions in India
mWellcare[28]	3698	mHealth system for integrated NCD management	Blood pressure and HbA1c control	No significant difference in HbA1c compared to enhanced usual care
K-DPP[37]	1007	Peer-supported lifestyle modification	Diabetes incidence	No significant reduction in diabetes incidence; Improved cardiovascular risk factors in rural communities
Chunampet project[31]	23380	Telemedicine with mobile screening van	HbA1c control	Mean HbA1c decrease from 9.3% to 8.5% (non-randomized design)

BP: Blood pressure; LDL: Low-density lipoprotein; NCD: Non-communicable disease.

Table 3 Key lessons from digital health tools applied for diabetes in rural China and India

Theme	Key lessons	Supporting evidence and examples
Health system context alignment	Digital interventions must align with existing healthcare structures and governance systems	SimCard tailored implementation to different health system contexts in China and India[9]; ROADMAP adapted to China's hierarchical healthcare system[8]
Targeting high-need populations	Digital interventions show greater effectiveness in populations with poorer baseline control and in remote areas	ROADMAP showed stronger effects for patients with baseline HbA1c > 8%[8]; Chunampet project demonstrated substantial HbA1c reduction in remote populations in India[31]; SMARTDiabetes was more effective in rural than urban areas in China[12]
Leveraging existing social structures	Family and community support structures are particularly effective in rural settings	Family Health Promoters assisting patients with self-management and use of the SMARTDiabetes app were associated with improved diabetes control in China[12]; Kerala Diabetes Prevention Program utilized trained peer leaders effectively in India[37]
Support task-sharing	Digital tools enable task-sharing among health professionals to reduce physician’s high workload	Dedicated digital platforms were developed for volunteer community health workers and licensed physicians who shared healthcare tasks in rural India in the SimCard trial[9]; India’s I-TREC model redistributed workflow from physicians to nurses, enabling nurses to conduct initial assessments[32]
Co-creation and stakeholder engagement	Co-creation enhances intervention acceptability, feasibility, and implementation potential	Jindal et al[32] developed I-TREC through multiple stakeholder advisory boards and formal partnerships with government agencies; Yin et al[33] systematically identified implementation barriers through stakeholder interviews before designing interventions in rural China