- Atlas
- Medicine & Health
- Digital Health
Digital Health
Digital health is the use of computing and communication technologies—records, sensors, phones, networks, algorithms—to support health and health care.
12 frameworks, from 1959 to today · Health Informatics · Medicine & Health
- Telehealth 1959-Present
- Clinical Information Systems and Electronic Health Records 1965-Present
- Problem-Oriented Medical Record 1968-1990
- Hospital information system 1970-1995
- Clinical decision support system 1975-Present
- Health IT Evaluation and Outcomes Research 1980-Present
- Electronic Health Record 1985-Present
- Health Information Exchange and Interoperability Standards 1987-Present
- Consumer Health Informatics 1995-Present
- National Health IT Policy and Meaningful Use 2004-Present
- MHealth 2005-Present
- Clinical Data Science and Machine Learning 2012-Present
Overview
Digital health is the use of computing and communication technologies—records, sensors, phones, networks, algorithms—to support health and health care1. Its practitioners build and study the systems that hold a patient's medical history, move data between hospitals and clinics, deliver care at a distance, and turn raw measurements into advice for clinicians and patients1. What is at stake is not gadgetry for its own sake: a patient's health is documented in many places, and the value of that information depends on whether it can be captured reliably, moved intact, and turned into a decision at the right moment. The history of digital health is the story of how the field learned to make that happen—and of the recurring gap between what systems promise and what they deliver.
Paper Records and the First Clinical Computers
The prehistory of digital health lies in the hospital record room. In the early twentieth century, hospitals professionalized medical records: standardized forms, trained record librarians, and the patient chart as a legal and clinical document. The problem this solved was continuity—a patient's history could follow them between wards and visits. The problem it created was scale: paper charts were bulky, hard to search, and easy to lose.
The first electronic medical records were built in the 1960s and 1970s, mostly in academic medical centers. At the University of Vermont, Lawrence Weed developed the problem-oriented medical record in the late 1960s, organizing the chart around a numbered list of the patient's problems rather than around chronological notes. Weed's insight was that a structured record could support reasoning, not just storage. Around the same time, systems such as the HELP system at LDS Hospital in Salt Lake City and the Regenstrief Medical Record System in Indianapolis began storing clinical data electronically and using it to generate reminders and reports. These were expensive, institution-specific systems, but they demonstrated that a computer could hold a patient's data and act on it.
Clinical Information Systems and the Electronic Health Record
The electronic health record (EHR)—a digital version of a patient's chart shared across care settings—became the central artifact of the field. Early hospital information systems in the 1970s and 1980s combined billing, laboratory, and pharmacy data; the clinical record was often a byproduct. The turning point came when institutions began building records designed for clinicians rather than administrators. The Department of Veterans Affairs' Decentralized Hospital Computer Program, begun in the late 1970s, grew into the Veterans Health Information Systems and Technology Architecture (VistA), a comprehensive EHR that by the 1990s was used across the VA system. In the 1980s and 1990s, commercial systems such as those from Epic and Cerner spread through large hospitals, and the term "electronic medical record" gave way to "electronic health record" as the ambition shifted from a single institution's chart to a longitudinal record across providers.
The persistent problem was adoption. Through the 1990s and early 2000s, most U.S. hospitals and clinics still ran largely on paper, and EHRs were criticized as costly, hard to use, and disruptive to clinical workflow. The field's answer was not only better software but policy.
National Health IT Policy and the Push for Adoption
In the United States, the Health Information Technology for Economic and Clinical Health (HITECH) Act of 2009 committed tens of billions of dollars in incentive payments to hospitals and clinicians who adopted certified EHRs and demonstrated "meaningful use"—using the systems in ways that improved care, such as electronic prescribing and reporting quality measures. Adoption rose sharply: by the mid-2010s, most U.S. hospitals and a majority of office-based physicians had adopted EHRs. Similar national programs appeared elsewhere, from the United Kingdom's National Programme for IT in the 2000s to Denmark's long-standing national health IT infrastructure. The policy turn made the EHR a near-universal tool, but it also exposed new problems: usability complaints, documentation burden, and the difficulty of making systems talk to one another.
Health Information Exchange and Interoperability Standards
The interoperability problem—getting data from one system to another without loss or distortion—was recognized early. In the late 1980s, a group of vendors and users formed Health Level Seven (HL7) to develop messaging standards; HL7 version 2, released in the late 1980s, became the most widely used health data messaging standard in the world, still carrying much hospital data today. Later efforts, including HL7 version 3 and the Clinical Document Architecture, aimed at richer, more structured exchange, and the Fast Healthcare Interoperability Resources (FHIR) standard, first published in the 2010s, offered a web-based, modular approach that has been widely adopted. Alongside messaging standards came terminology systems—SNOMED CT for clinical concepts, LOINC for laboratory observations—that let different systems agree on what a piece of data means23. Health information exchange organizations, regional and national, tried to knit these pieces into working networks4. The results were uneven: standards made exchange possible, but organizational, legal, and economic barriers often made it rare.
Clinical Decision Support Systems
If the EHR is the record, clinical decision support (CDS) is the record put to work. CDS systems deliver reminders, alerts, order sets, and reference information at the point of care5. Early examples include the HELP system's alerts and the Regenstrief system's reminders in the 1970s and 1980s; a landmark randomized trial published in 1994 by David Bates and colleagues at Brigham and Women's Hospital showed that computer-generated alerts could reduce serious medication errors. The field learned that CDS works best when it is specific, timely, and integrated into workflow—and that poorly designed alerts cause "alert fatigue," training clinicians to ignore them. Later work extended CDS to diagnostic support, risk prediction, and guideline implementation, and the same logic now underlies much of the machine-learning wave.
Evaluation and Outcomes Research
From early on, digital health practitioners insisted that systems be evaluated, not just deployed. The field developed a tradition of health IT evaluation: randomized trials of reminders and alerts, before-and-after studies of EHR implementation, and analyses of unintended consequences. A widely cited 2005 paper by Jonathan Nebeker and colleagues documented ways computer systems could introduce new errors—wrong-patient selection, data-entry mistakes, confusing displays. Evaluation research gave the field a vocabulary for asking whether a system actually improved care, and it repeatedly found that the answer depended on context: the same software could help in one setting and harm in another. This skepticism became a defining feature of the discipline.
Consumer Health Informatics and Telehealth
Digital health also moved outside the clinic. Telemedicine—care delivered at a distance—has roots in mid-twentieth-century experiments: clinicians at the Nebraska Psychiatric Institute used closed-circuit television for consultations in the 1950s, and NASA and the Indian Health Service ran early telemedicine projects in the 1960s and 1970s. The internet transformed the scale. In the 1990s, consumer health informatics emerged as people began seeking health information online, and researchers studied the quality of that information, the design of decision aids, and the rise of online patient communities. The term "e-patient"—empowered, engaged, equipped, enabled—captured a shift toward patients as active users of health information. Telehealth expanded from video consultations to remote monitoring and store-and-forward messaging, and the COVID-19 pandemic of 2020 forced a sudden, massive shift to virtual care that made telehealth a routine part of practice.
Mobile Health and Wearable Sensing
The smartphone put a computer in every pocket. Mobile health (mHealth) grew through the 2000s and 2010s: apps for medication adherence, symptom tracking, and diabetes management; text-message reminders in low-resource settings; and wearable sensors measuring heart rate, activity, and sleep. The field's challenge shifted from capturing data to making sense of it—separating signal from noise, integrating consumer devices with clinical systems, and addressing privacy and equity. Wearables and remote monitoring promised continuous, real-world data, but questions about accuracy, regulation, and whether such data improve outcomes remained open.
Clinical Data Science and Machine Learning
The most recent turning point is the application of machine learning—statistical methods that learn patterns from data rather than following explicit rules—to clinical data. Early clinical prediction models, such as the APACHE score for intensive-care mortality in the 1980s, used regression; the 2010s brought deep learning and large-scale electronic health record datasets. Landmark results included deep-learning models for diabetic retinopathy screening and sepsis prediction, and the development of large language models raised new possibilities for documentation and decision support. The field's evaluation tradition reasserted itself: concerns about algorithmic bias, generalization across institutions, and the gap between prediction and improved outcomes became central research questions.
The Present
Today digital health is a mature, sprawling field. The EHR is near-universal in wealthy health systems, interoperability is advancing through FHIR and national exchange networks, and clinical decision support is increasingly powered by machine learning. Telehealth and mobile sensing are routine parts of care, and consumer tools blur the line between patient and clinician. The open questions are the ones the field has carried from the beginning: how to make systems usable and trustworthy, how to move data without losing meaning, how to evaluate whether technology actually improves health, and how to ensure that the benefits reach everyone. The history of digital health is a history of learning that the hard part is rarely the computer.
References
- Digital health, Wikipedia
- SNOMED CT, Wikipedia
- LOINC, Wikipedia
- Health information exchange, Wikipedia
- Clinical decision support system, Wikipedia
The frameworks of Digital Health
Each framework is a way of working the field has used: when it held, and the question it answers.
- 1959-Present
- 1965-Present
Clinical Information Systems and Electronic Health Records
How should clinical data be stored and used?
- 1968-1990
Problem-Oriented Medical Record
How should a patient's chart be organized?
- 1970-1995
Hospital information system
How can a hospital's operations be computerized?
- 1975-Present
Clinical decision support system
How can computers help clinicians decide?
- 1980-Present
Health IT Evaluation and Outcomes Research
Does health IT actually improve care?
- 1985-Present
Electronic Health Record
How can a patient's record follow them everywhere?
- 1987-Present
Health Information Exchange and Interoperability Standards
How can health data move between systems intact?
- 1995-Present
Consumer Health Informatics
How should patients use health information themselves?
- 2004-Present
National Health IT Policy and Meaningful Use
How can policy make health IT universal?
- 2005-Present
- 2012-Present
Clinical Data Science and Machine Learning
Can machines learn to predict health outcomes?
How the ideas connect
Each arrow runs from an earlier idea to the later one it shaped. The list below says what passed between them.
Clinical decision support system influenced by Clinical Information Systems and Electronic Health Records
Clinical decision support grew directly out of clinical information systems: the HELP and Regenstrief systems stored clinical data electronically and used it to generate reminders and alerts, establishing the pattern CDS later generalized.
Carried by HELP system; Regenstrief Medical Record System.
National Health IT Policy and Meaningful Use influenced by Electronic Health Record
The HITECH Act's meaningful-use incentives were built around adoption of certified EHRs, making the electronic health record the target of national policy and tying payment to its use.
Carried by HITECH Act (2009).
Health Information Exchange and Interoperability Standards influenced by Clinical Information Systems and Electronic Health Records
The spread of heterogeneous clinical information systems created the practical need for messaging standards, and vendors and users formed HL7 in the late 1980s to move data between them.
Carried by Health Level Seven.
Clinical Information Systems and Electronic Health Records influenced by Problem-Oriented Medical Record
Weed's problem-oriented record supplied the organizing concept of the problem list and structured clinical data that early clinical information systems adopted as the backbone of the electronic chart.
Carried by Lawrence Weed.
Health IT Evaluation and Outcomes Research influenced by Clinical decision support system
Randomized trials of computer-generated alerts, notably the 1994 Brigham and Women's study of medication-error reduction, gave health IT evaluation its core method and subject matter.
Carried by David Bates.
Clinical Data Science and Machine Learning influenced by Clinical decision support system
The logic of point-of-care reminders, risk prediction and guideline implementation carried into machine-learning clinical data science, which recast decision support as learned prediction from data.
Health Information Exchange and Interoperability Standards influenced by Electronic Health Record
The ambition of a longitudinal record across providers made interoperability a central requirement, driving later standards work such as the Clinical Document Architecture and FHIR.
Clinical Data Science and Machine Learning influenced by Electronic Health Record
The accumulated digital record data in EHRs became the training material for clinical machine learning, which depends on the structured and unstructured data EHRs capture.
National Health IT Policy and Meaningful Use influenced by Health Information Exchange and Interoperability Standards
Meaningful-use criteria required electronic exchange and reporting of quality measures, embedding interoperability standards in national health IT policy.
MHealth grew out of Consumer Health Informatics
Mobile health extended consumer health informatics from web-based information seeking and decision aids to smartphone apps and wearable sensors used by patients directly.
MHealth influenced by Telehealth
Telehealth's remote monitoring and messaging models carried over into mHealth as smartphones became the delivery platform for distance care and adherence support.
Beyond the field
Where Digital Health's ideas came from in other fields, and what they led to there: the strongest 12 of 87 links.
Telehealth led to Consumer Telehealth and Remote Monitoring in Consumer Health Informatics
Consumer telehealth and remote monitoring is the extension of telehealth from clinician-to-clinician consultation to direct patient contact and home devices. The store-and-forward and real-time telemedicine techniques developed earlier were repurposed for patients managing their own conditions at a distance.
Clinical Information Systems and Electronic Health Records influenced by Problem-Oriented Medical Record in Clinical Informatics
The problem list and SOAP note structure of the problem-oriented record became a standard component of electronic health record design.
Carried by Lawrence L. Weed.
Problem-Oriented Medical Record influenced Computer-Based Patient Record in Clinical Informatics
The problem-oriented medical record's problem list and structured progress notes became the organising scheme of computer-based patient record systems.
Carried by Lawrence L. Weed.
Hospital information system led to Electronic health record in Clinical Informatics
Hospital information systems of the 1970s to 1990s supplied the administrative and clinical data infrastructure from which the electronic health record was assembled. The EHR extended the hospital system's record-keeping to a longitudinal patient-centred record across settings.
Health IT Evaluation and Outcomes Research grew out of Outcomes Research in Comparative Effectiveness Research
Health IT evaluation and outcomes research grew out of outcomes research, adopting its methods for measuring the effects of care on patient outcomes and applying them to information technology interventions. The outcomes framework was extended to evaluate health IT.
Electronic Health Record grew out of Computer-Based Patient Record in Clinical Informatics
The computer-based patient record, codified in the Institute of Medicine's 1991 report, was the direct precursor of the electronic health record. The later EHR extended the CPR from an episode- or clinic-centred record to a longitudinal record shared across institutions.
Carried by Institute of Medicine; The Computer-Based Patient Record: An Essential Technology for Health Care (1991).
Clinical Data Science and Machine Learning influenced Predictive Analytics and Algorithmic Governance in Nursing Informatics
Clinical data science and machine learning in digital health provided the algorithms and data pipelines that the nursing predictive analytics and governance framework regulates. The framework's governance questions follow directly from these models' clinical deployment.
Telehealth led to EHealth and mHealth in Health Communication
eHealth and mHealth grew out of telehealth's use of telecommunications to deliver care at a distance, extending it from clinical consultation to internet and mobile consumer-facing health services.
Clinical Information Systems and Electronic Health Records influenced by Clinical Classification and Terminology Systems in Health Informatics
Electronic health record systems had to encode diagnoses and procedures, so they adopted the coded vocabularies and classification schemes built for clinical and statistical use. Standard terminologies became the semantic backbone of the record.
Clinical Information Systems and Electronic Health Records influenced by Hospital Standardization and Medical Record Practice in Health Data Standards
Hospital standardization and medical record practice established the standards for complete, uniform, retrievable patient records that electronic systems then implemented. The accreditation-driven record requirements shaped the data elements and documentation rules of the EHR.
Clinical Information Systems and Electronic Health Records influenced by Paper record and Kardex systems in Nursing Informatics
Early electronic record designs took the structure and workflow of paper charts and nursing Kardex systems as their template, automating the same forms and order of entries. The paper record defined what the electronic version had to reproduce.
Clinical Information Systems and Electronic Health Records grew out of Problem-Oriented Medical Record in Health Informatics
Clinical information systems and electronic health records adopted the problem-oriented record's problem list and SOAP note structure as core components of the electronic record.
Carried by Lawrence Weed.

Digital Health is a planet in Nousfarer
Its city has a tower for each of these 12 frameworks. Land, walk in, and climb one concept at a time.
Land on Digital HealthRelated fields
Fields its ideas are linked to
- Health Informatics 16 links
- Clinical Informatics 14 links
- Health Information Systems 13 links
- Imaging Informatics 9 links
- Consumer Health Informatics 8 links
- Health Data Standards 7 links
- Public Health Informatics 3 links
- Health Promotion 3 links
- Nursing Informatics 3 links
- Health Communication 2 links
- Health Policy Evaluation 2 links
- Biomedical Engineering 1 link