Page 17 - AIH-1-3

Artificial Intelligence in Health Optimizing EHRs to support AI

problems and/or to consider how best to achieve desired reliably support AI technologies. This paper explores
health outcomes. what our future infrastructure needs are to ensure
AI is used as an aid to problem solving. When used to flexibility and agility to adapt to new requirements and
support clinical decisions, AI cannot operate like a “black- shifting priorities. It is about informing politicians, top
box.” For AI to be trusted by health-care professionals, public officials, senior decision-makers, and procurement
these insights, inferences, and decision supports need to be officers as well as data scientists working for the health
explainable and repeatable. All types of data/information industry.
can be made use of by AI. In the health-care domain, all 1.1. Review aims and objectives
relevant legislative, regulatory requirements, and ethical
principles need to be complied with for obtaining the The aims and objectives of this review are as follows:
authorization to use of these resources. • To identify the current state of data availability, and
quality for AI development use.
Many of AI data resources come from electronic • To identify foundational data/information AI resource
health records (EHRs) and electronic medical records needs.
(EMRs). Based on the International Organization for • To explore the benefits of new and emerging technologies
Standardization (ISO) Technical Committee (TC) 215 used to design and implement next-generation EHR/
definitions, an EHR is a health record in computer- EMRs.
processable format – such records have atomic data • To identify new-generation capacity to optimize AI
elements, representing the smallest possible concept such developments.
as genetic data, which can be analyzed and processed by a
computer. Medical (healthcare) records are health records 1.2. Research question
produced for and used within a health-care organization’s What next-generation technologies need to be in place to
enterprise system. Many health-care organizations begin optimize EHRs for AI purposes?
their digital health transformation journey by scanning
paper-based records. The content of EHR needs to enable 1.3. Rapid review method
continuous monitoring of people’s health throughout The scope of this modified scoping (rapid) review paper
a person’s life span, facilitate continuity of care, and is confined to the use of data resources as stored for use
be informative and accessible to them anytime from by current and next-generation EHRs/EMR systems for AI
anywhere, as well as accessible to others authorized to do purposes. This review was designed to benefit high-level
so with consent. decision-makers who need a conceptual understanding
EMRs represent records produced for and used by one of current issues and how these are best addressed. The
health-care provider (independent or organizational). selection of references used was primarily informed by
A growing number of EMR suppliers are providing access to the authors’ expert knowledge and shared experiences
other providers as well as their patients through a dedicated between standards development experts and known
portal. Health information exchange (HIE) refers to the researchers working in the digital health space. This review
concept of exchanging clinical and administrative data was supplemented by targeted research methods based on
across different systems and stakeholders, by connecting literary information sources, including the most up-to-
the platform to transactional systems. In most cases, EHR date gray literature and known published literature not
data can be gleaned from a multitude of proprietary EMR retrieved by database searches.
and HIE systems as well as a myriad of other clinical,
administrative, and ancillary systems. This current 2. Issues associated with the current state
fragmented landscape of health information systems’ data of health data
acquisition, storage, and use is a barrier to the effective Individual health-care organizations began the development
provision of continuous and person-centered care as well of EHR/EMR systems alongside their evolving computer
as trustworthy use of AI methods. science, information and communication technologies,
As clinicians, our focus is on ensuring that health data and information systems research and development
used for AI purposes are complete, accurate, trustworthy, activities. Some of these early developed systems have
and able to support person-centered precision medicine. evolved over more than 30 years by taking advantage of new
This requires a major overhaul of this fragmented insights, newly discovered technologies, new programming
landscape by facilitating collaboration among diverse languages, and frameworks and evolving connectivity
health service providers and across different types of care. solutions, to become some of the mega-EMR enterprise
Only then are we able to generate timely quality data to systems still in use today. These are now legacy systems

Volume 1 Issue 3 (2024) 11 doi: 10.36922/aih.3056