What is digital health?

Digital health encompasses the use of digital technologies — information and communication technologies — to improve health outcomes and health service delivery. Key domains: Telemedicine: remote clinical consultations via video, telephone or asynchronous messaging. mHealth: mobile phone apps and SMS for health information, monitoring and behaviour change. Artificial intelligence (AI): machine learning algorithms for diagnosis, clinical decision support, drug discovery. Electronic health records (EHR): digital patient records replacing paper. Wearable devices and remote monitoring: smartwatches measuring heart rate, ECG, oxygen saturation; continuous glucose monitoring. Digital disease surveillance: real-time monitoring of disease patterns for outbreak detection (WHO EpiPulse, ProMED, DHIS2).

What is the role of AI in medicine?

AI is transforming multiple areas of medicine: Diagnostics: AI algorithms achieving specialist-level accuracy in chest X-ray interpretation, diabetic retinopathy screening, dermatology, pathology and genomics — potentially extending specialist access to resource-limited settings. Drug discovery: AlphaFold protein structure prediction; generative AI designing novel drug molecules. Clinical decision support: AI analysing EHR data to predict sepsis, deterioration, medication errors. Administrative AI: automated coding, scheduling, documentation. Limitations: AI performs best on well-defined, image-based tasks; generalisability across different populations is not guaranteed; most AI tools validated on Western, well-resourced population data.

How did COVID-19 transform digital health?

The COVID-19 pandemic was a catalyst that accelerated digital health adoption by an estimated 5-10 years: Telemedicine: routine outpatient consultations shifted to video (telehealth encounters increased 60× in the US in 2020); reimbursement frameworks changed permanently. Digital contact tracing: Bluetooth-based exposure notification apps (COVID Safe, StopCovid) deployed across 50+ countries. Digital vaccine certificates: QR code-based COVID vaccination records deployed globally. Data sharing: unprecedented real-time genomic and epidemiological data sharing via GISAID, WHO. Mental health apps: dramatic growth in digital mental health platforms.

What are the risks of AI in healthcare?

Key risks identified by WHO (2021): Bias and inequity: AI trained on biased datasets (predominantly Western, White, male, urban populations) can perform worse for underrepresented groups — potentially widening health disparities. Automation bias: clinicians over-relying on AI recommendations, not applying critical judgment. Liability: unclear accountability when AI causes harm — who is responsible? The clinician? The AI developer? The hospital? Data security: medical AI systems can be vulnerable to adversarial attacks (deliberately manipulated inputs causing wrong outputs). Digital divide: AI benefits accruing to wealthy, connected populations rather than those who need them most.

How does GMJ cover digital health?

GMJ News covers WHO's Global Strategy on Digital Health 2020-2025, AI diagnostic tools achieving specialist accuracy, telemedicine evidence and access, AlphaFold and protein structure revolution (Nobel 2024), mHealth in LMICs, WHO AI ethics framework, GDPR health data governance, and digital health equity challenges. Digital health is directly relevant to PHIG's health information systems, Georgia's health digitisation agenda, and GMJ's role in evidence-based health communication.

Home › Topics › Digital Health

Digital Health

GMJ News knowledge hub · last reviewed September 2026 · Georgian Medical Journal

Summary Statistics Glossary GMJ news FAQ Documents Organizations Research

Digital health — encompassing telemedicine, mobile health (mHealth), artificial intelligence (AI) in clinical decision-making, electronic health records (EHR), wearable sensors and genomics data platforms — has become a defining priority of twenty-first century healthcare, with WHO publishing its Global Strategy on Digital Health 2020-2025 and the world witnessing AI diagnostic tools achieve specialist-level accuracy in radiology, ophthalmology (diabetic retinopathy screening — WHO-approved AI tools), pathology and dermatology (WHO). The post-COVID telemedicine revolution brought 150 million+ new telehealth encounters in 2020 alone (US McKinsey data) while AlphaFold (DeepMind, 2021-2022) predicted the structure of virtually all known proteins — transforming drug discovery — and the WHO’s 2021 AI Ethics and Governance guidelines established the six principles (transparency, inclusiveness, responsibility, impartiality, security, accountability) that frame responsible AI deployment in health.

Key messages

WHO Global Strategy 2020-2025 — digital health as health equity tool

WHO's Global Strategy on Digital Health 2020-2025 positions digital health not as a luxury but as an essential equity tool — enabling universal health coverage through telemedicine, mHealth, AI diagnostic tools and digital disease surveillance, particularly in under-resourced settings (WHO 2020).

AI achieving specialist-level accuracy in diagnostics

Artificial intelligence diagnostic tools are achieving specialist-level performance in radiology (chest X-ray, CT, MRI interpretation), ophthalmology (diabetic retinopathy screening — WHO-approved AI tools), dermatology (skin lesion classification), pathology and genomics — with implications for extending specialist capacity to resource-limited settings.

Telemedicine — 150M+ new encounters in 2020

The COVID-19 pandemic accelerated telemedicine adoption by approximately 10 years. Telehealth encounters increased from approximately 840,000 (2019) to 52+ million (2020) in the US alone. Telemedicine improves access for rural populations, elderly housebound patients and those with mobility limitations.

AlphaFold — the protein structure revolution

DeepMind's AlphaFold (2021-2022) predicted the 3D structures of virtually all known proteins (approximately 200 million structures) — solving a 50-year grand challenge in biology and transforming drug discovery, vaccine design and fundamental biological research. Nobel Prize in Chemistry 2024 awarded to AlphaFold creators.

Data privacy — health data is most sensitive personal data

Health data is the most sensitive category of personal data under GDPR and equivalent legislation. Digitisation of health records creates vast opportunities for research but requires robust governance: patient consent, data minimisation, purpose limitation, pseudonymisation and security. The WHO's 2021 AI Ethics guidelines establish six principles for responsible health AI.

Digital divide — risk of widening health inequalities

Digital health technologies risk exacerbating health inequalities if they predominantly reach tech-literate, connected, urban populations — while bypassing the elderly, low-income, rural and digitally excluded. WHO and international bodies call for digital health to be designed with equity as a core principle, not an afterthought.

Key statistics

2020-2025

WHO Global Strategy on Digital Health — the governing framework

WHO 2020

200M

protein structures predicted by AlphaFold 2021-2022 (DeepMind)

DeepMind/Nature 2021

52M+

telehealth visits in US in 2020 (vs 840K in 2019) — 60× increase

McKinsey/HHS

principles in WHO AI Ethics and Governance of AI for Health (2021)

WHO 2021

Nobel 2024

Nobel Prize in Chemistry awarded to AlphaFold protein structure prediction

Nobel Committee 2024

SMART

WHO guidelines (Standardised, Machine-readable, Adaptive, Requirements-based, Testable) for digital health

WHO SMART 2021

Digital health ecosystem — key domains and WHO priority areas

Source: WHO Global Strategy on Digital Health 2020-2025. All domains contribute to universal health coverage.

Glossary of key terms

WHO Global Strategy on Digital Health
WHO 2020Adopted by the 73rd World Health Assembly in 2020. Four strategic objectives: promote global collaboration on digital health; advance appropriate implementation at country level; strengthen governance for digital health; establish the evidence base through research and evaluation. WHO SMART guidelines (Standardised, Machine-readable, Adaptive, Requirements-based, Testable) provide the operational framework for implementing digital health in clinical and public health settings.

AI in diagnostic imaging
WHO/FDAAI algorithms trained on large datasets of medical images can detect abnormalities with specialist-level accuracy: Chest X-ray: tuberculosis detection, pneumonia classification, COVID-19 patterns. Ophthalmology: diabetic retinopathy screening (WHO-approved AI tools deployed in LMICs); glaucoma detection; AMD. Dermatology: melanoma vs benign lesion classification — AI achieving dermatologist-level accuracy. Radiology: lung nodule detection, breast cancer mammography AI — FDA-approved tools. Pathology: cancer grade classification from histology slides.

AlphaFold (DeepMind)
DeepMind/Nature 2021An AI system using deep learning to predict 3D protein structures from amino acid sequences with unprecedented accuracy — solving the 50-year protein folding problem. Published July 2021 (Nature). Version 2 predicted structures for essentially all known human proteins and 98.5% of structures in the UniProt database (approximately 214 million proteins). Impact: drug discovery acceleration; vaccine design; understanding genetic disease mechanisms; enzyme design for biosynthesis. Nobel Prize in Chemistry 2024: David Baker, John Jumper, Demis Hassabis.

mHealth (mobile health)
WHO/GSMAHealth services and information delivered through mobile technologies — particularly important in LMICs with high mobile phone penetration but limited healthcare infrastructure: SMS medication adherence reminders (shown to improve HIV ART adherence); DHIS2 (District Health Information System 2) — the most widely used health data management system in LMICs; community health worker digital tools; maternal health apps; outbreak surveillance (WHO Go.Data, DHIS2 tracker).

WHO AI Ethics: 6 principles
WHO 20211. Transparency: clear documentation of AI decision-making. 2. Inclusiveness: AI trained on diverse, representative data; does not perpetuate bias. 3. Responsibility: clear accountability for AI decisions; human oversight. 4. Impartiality: equitable access; not only for wealthy or technically sophisticated. 5. Security: robust data protection, resistance to adversarial attack. 6. Accountability: clear audit trails; redress mechanisms. Published in WHO's Ethics and Governance of AI for Health (2021).

GDPR and health data
EU/WHOHealth data is a "special category" of personal data under EU GDPR (Article 9) — requiring explicit consent or other specific lawful basis for processing. Key principles: data minimisation (only collect what's needed); purpose limitation (not use beyond original purpose); pseudonymisation (replace identifying data with pseudonyms); data subject rights (access, rectification, erasure). Non-GDPR countries: WHO advocates for analogous national legislation governing health data. Secondary use of health data for research: increasingly enabled by trusted research environments (TREs) and federated learning.