Microsoft Word - IDPT 2007 INTEROP Paper 5 Begoyan.doc

Integrated Design and Process Technology, IDPT-2007

Printed in the United States of America, June, 2007

2007 Society for Design and Process Science

AN OVERVIEW OF INTEROPERABILITY STANDARDS FOR

ELECTRONIC HEALTH RECORDS

A. Begoyan

University of Westminster

School of Informatics

Department of Information Systems

115 New Cavendish Street, London, W1W 6UW

ABSTRACT

This paper gives an overview of the interoperability

standards for Electronic Healthcare Records (EHRs)

which are offered by the world leading standard bodies:

European Committee for Standardisation (CEN),

International Organization for Standardization (ISO),

Health Level Seven (HL7) and Digital Imaging and

Communications in Medicine (DICOM). We have taken

information available in the original standards’ text as the

bases of our research. We outline what each of these

standards offer and we compare the way they address the

interoperability of EHRs.

INTRODUCTION

Healthcare information systems have grown rapidly

in the last decade. They moved from isolated software

systems in hospitals or primary care organizations,

towards solutions which support a continuous medical

process and (i) include multiple healthcare professionals

and institutions, (ii) utilize ubiquitous computing

healthcare environments and (iii) embrace technological

advances, typical of the domain of today’s pervasive

software applications. It is not a surprise that in such an

interactive environment, we need to look at the

information sharing amongst healthcare software systems.

This requires that all relevant patient clinical data is

available and correct, thus supporting healthcare

professionals at any time, any place and improving the

quality and delivery of healthcare. We have witnessed a

proliferation of solutions for effective EHR (Dumay,

2002), (van Ginneken 2002), (Giusse, 2003), (Ueckert,

2003), (Valdes, 2004), (Kuhn, 2001), (Hsyien-Chi, 2007).

We are now in a specific situation characterized by:

(a) Heterogeneities in hardware and software

solutions that support healthcare software applications as

a consequence of their evolution from the early 1960s;

(b) Heterogeneities in the structure, purpose and

deployment of EHR in terms of their suitability for a

variety of healthcare systems that are growing in each

country in the world.

Heterogeneities in healthcare software applications

from (a) above are typical of heterogeneities in any other

problem domain, where a variety of platforms, systems,

databases, software applications, technologies etc exist.

From this point of view, healthcare software systems do

not differ from any other. However, heterogeneity of

EHRs across healthcare systems from (b) above is a

bigger problem. EHRs differ from application to

application and from country to country. This means that

the structure of EHRs and the methods used for

exchanging their content may vary significantly, thus

becoming an obstacle for sharing medical data or medical

records and developing software applications that support

(i)-(iii) above.

The expected answer is in the standardization of

EHRs structure, content and the way of exchanging them,

as described in (Gross, 2005), (Kalra, 2006), (Bossen,

2006). There are initiatives from North America and

European countries to engage their organizations and

bodies responsible for standardization in general, with the

EHR standardization. Thus we have various EHR

standards which differ from country to country and which

cover different issues of EHR standardization. On top of

that, these standards are not very stable. There are a

number of reasons for that. Standards are constantly

evolving and changing, and the content and structure of

EHRs is also changing in parallel due to technology and

science advancement, thus pushing changes within

applications built upon such EHRs outside their initial

domains and purposes (Beale, 2002). Even if we have

achieved a certain level of EHR standardization, the

advancement of medial science would demand further

changes in existing EHR standards. For example, medical

science is currently placing great emphasis on genomics,

which requires the integration of biomedical information

with the EHR. If the current EHR structure has not

already incorporated biomedical information it will have

to be done soon. Therefore, standardization of the EHR

needs to be reviewed on a regular basis to asses the

current situation within the EHR standardization, and take

steps to make EHR standards ‘future proof’ or adaptable

to inevitable changes (Medisell, 2004).

Healthcare informatics is full of various standards

that help us to structure many aspects of healthcare

software systems: from their architectures, software

applications built from them to semantics stored within

and structures of EHR. In this paper we overview

standards for the interoperability of EHRs which exist

across North America and Europe. We are interested in a

particular aspect of interoperability between diverse

healthcare systems, which should be able to exchange

information about patients and their medical history stored

within EHRs (Bott, 2004). Therefore:

We are interested in software engineering solutions

that can alleviate interoperability problems in healthcare

software applications, which depend on the data structures

of EHRs. Thus any interoperability standard on EHR

could be a starting point when building such applications.

We have learned that in spite of various initiatives for

achieving the interoperability in healthcare software

applications through data sharing, such as e-GIF (Cabinet

Office, 2005) in the UK, current EHRs are simply not

equipped with data items, which are essential if we want

to share them across healthcare institutions (Slevin et al.,

2005) (Akram et al., 2007) .

Five year ago the UK government initiated the

National Program for Information Technology (NPfIT,

2004), which has already approved the development of a

centralized healthcare (patient) database for the UK

National Heath Service. Which standards on EHRs would

they follow, if any?

In order to review the situation within standards,

which deals with the interoperability of EHRs, we started

with the European standardization committee and its

Interoperability working group IV within Technical

Committee 251 (CEN/TC 251). They list the following

standardization bodies responsible for EHR.

ISO is an International Organization for

Standardization which is currently a network of the

national standards institutes of 157 countries. They work

on the basis of one member per country, with a Central

Secretariat in Geneva, Switzerland, that coordinates the

system. ISO produces EHR standards that are limited to

the structure and function of the EHR and the system that

processes EHR.

CEN is a European Committee for Standardisation. It

is involved in developing multi-disciplinary standards

including health care systems and their interoperability.

CEN covers European Union (EU) countries and some

affiliated countries outside the EU.

HL7 stands for Health Level Seven. It is one of the

several American National Standards Institute (ANSI) -

accredited Standards Developing Organizations, which

operates in the healthcare arena. It covers America, some

European and Asian countries and Australia. Its purpose is

to provide standards for data exchange between different

types of healthcare computer applications. HL7’s domain

includes clinical and administrative data. Its headquarters

is in Ann Arbor, Michigan.

DICOM – Digital Imaging and Communications in

Medicine is an association of medical industry and

medical professional organizations, working under the

umbrella of the National Electrical Manufacturers

Association (NEMA). They have delivered DICOM,

which is a de facto standard for medical image

communication (DICOM). This standard allows an

independent exchange between medical images and

related information.

The purpose of this paper is to briefly overview the

current situation within standards that regulate the

interoperability of EHRs and to learn the outcomes of

such standardizations. These outcomes then can be

adopted when software applications that deal with

heterogeneities are designed. We would also like to see if

there are any components within such standardizations

which lead towards ‘future-proof’ EHRs.

This paper is structured as follows. In the next section

we define and discuss EHRs, their interoperability and

prerequisites for their semantic interoperability. In the

four sections that follow we discuss four EHR standards:

ISO/TR 20514, ISO/TS 18308, ISO/TR 18307 derived by

ISO; CEN13606 and its five parts from CEN; Version 2x

and Version3x from HL7; and two standards from

DICOM - Web Access to DICOM Persistent Objects and

DICOM Structured Reporting. We briefly touch the

harmonization between standards, and conclude in the last

section.

DEFINITIONS OF EHR AND INTEROPERABILITY

The definition of EHR varies from standard to

standard. According to CEN/TC251 Health Informatics,

EHR is “a healthcare record in a computer readable

format”. According to ISO/TR 20514 the basic-generic

EHR is “healthcare record in a computer processable

format, which encapsulates readability but extends this to

include the notion that information in the EHR must be

amenable to programmatic manipulation and therefore to

automatic processing” (ISO TC 215, ISO/TR 20514,

2005).

‘EHR systems’ are defined uniformly across CEN

and ISO: as “system[s] for recording, retrieving and

manipulating information in electronic health records”.

Interoperability of EHR as defined in ISO (ISO TC

215, ISO/TR 20514, 2005) is “the ability of two or more

applications being able to communicate in an effective

manner without compromising the content of the

transmitted EHR”. They claim that it is important to

develop national and international standards for EHR

interoperability to be able to: (1) share patient health

information between health professionals in a multi-

disciplinary shared-care environment: (2) share patient

health information between organizations within an

enterprise, a regional or national health system, or across

national borders and (3) support interoperability between

software from different vendors.

There are two types of interoperability that are

relevant in this context. The ISO Technical Report (ISO

TC 215, ISO/TR 20514, 2005) defines these as Functional

interoperability and Semantic interoperability. Functional

interoperability deals with the exchange of information

between two or more systems in a format that is readable

by humans. Semantic interoperability deals with the

exchange of information between systems in a format that

is computer processable by the receiving system. We

could not find within the Technical Report explicitly listed

prerequisites for achieving Semantic Interoperability.

Thus we have to refer to the conference paper (Schloeffel

P., 2004) which states that there are four prerequisites that

need to be fulfilled in order to achieve semantic

interoperability: a) a standardized EHR Reference Model.

This deals with the semantics of EHR information

structures; b) a standardized service interface. This deals

with the semantics of interfaces between EHR and other

services: c) a standardized set of domain-specific concept

models. This deals with archetypes and templates for

different domain concepts: d) standardized terminologies.

This defines the language that is used in archetypes. (The

first two prerequisites also apply to functional

interoperability.)

INTERNATIONAL STANDARDS ORGANISATION

ISO’s Technical Committee ISO/TC 215, which deals

with Health Informatics, is an international standards

body, whose aim is to achieve compatibility and

interoperability between independent healthcare systems.

This is a relatively new standards body, which has

produced a set of new standards as well as using other

international standards from HL7, DICOM and CEN as

the basis for their standards. ISO has already published 37

standards relating to different aspects in the field of

medical informatics across 22 countries, including the

UK. The standards produced by ISO/TC 215 are being

observed in 20 countries including Central America,

Europe, Asia and the Middle East.

The following three sub-sections will introduce three

standards – ISO/TR 20514, ISO/TS 18308 and ISO/TR

18307 which are concerned with the EHR interoperability.

ISO/TR 20514 - EHR Definition, Scope and Context.

ISO/TR 20514 is a standard that defines the content

of the EHR, its structure and the context in which the

EHR is used. It also gives definitions of the terminology

used. ISO/TR 20514 defines the EHR structure through

“basic-generic EHR” (ISO TC 215, ISO/TR 20514, 2005).

Thus, it allows the broadest applicability of the given

EHR structure to a wide range of existing and future types

of EHRs and EHR systems. This, in turn, provides

support for legislative and access control requirements

that would be applicable to all forms of EHR. The basic-

generic EHR definition is supplemented by the detailed

definition which is stated in 1 and 2 below:

1. The ability to share patient health information

between authorized users of the EHR.

2. The ability to supporting continuing, efficient

and quality integrated health care.

According to (ISO/TR 20514, 2005), the sharing of

EHR information can take place at three different levels:

LEVEL1 is between different clinical disciplines or other

users, who could be using the same application but require

different or ad hoc organization of EHRs. LEVEL2 is

between different applications at a single EHR node. (An

EHR node is a physical location where EHRs are stored

and maintained.) LEVEL3 is across different EHR nodes.

When EHRs are capable of being shared at LEVEL3

then the EHRs are called Integrated Care EHRs (ICEHR).

ISO/TS 18308. Requirements for an EHR Reference

Architecture.

This is a new standard that gives requirements for the

architecture of EHR systems and not the specification for

such architectures (ISO TC 215, ISO/TS 18308, 2004). It

specifies the assembling and collating of clinical and

technical requirements for EHR architecture to support

usage, sharing and exchanging of EHRs across different

countries, different health sectors and different models of

healthcare delivery. The main users of this standard will

be developers of EHR architecture standards such as CEN

EN13606 and other reference architectures such as

openEHR (Eichelberg et el, 2005), (openEHR).

ISO TR 18307 – Interoperability and Compatibility in

Messaging and Communication Standards.

This standard describes the main requirements for

achieving interoperability and compatibility in trusted

health information interchange between software

applications and systems in healthcare. The standard

specifies the interoperability needs of the healthcare

community for the subject of care, the healthcare

professional, the healthcare provider organization, its

business units and the incorporated delivery network. It

also provides a criterion for developers and implementers

of standards for messaging and communication in the

healthcare domain. It lays down the foundation for the

health information interchange to be as trustful as possible

(ISO TR 18307).

CEN

CEN’s Technical Committee CEN/TC 251 deals with

the production of Standards in Health Informatics. It has

produced the only comprehensive EHR interoperability

standard in the world called CEN 13606.

CEN13606

CEN 13606 has been previously published as a pre-

standard named ENV 13606 in 1999 (Kalra and Ingram,

2006). It was not as successfully implemented as

expected due to various weaknesses, such as the outdated

architectural concepts of applications that support EHRs,

that were recommended by the standard (Eichelberg et el,

2005). In 2001, CEN/TC updated ENV13606 and

adopted the openEHR “archetype methodology” defined

by the openEHR Foundation to make it a full European

Standard EHR (Schloeffel P., 2004), (The OpenEHR

Foundation). The result of this work is in CEN

prEN13606. CEN prEN13606 has been adopted by 48

countries as at 2007.

CEN EN13606 is a five-part standard as defined in

prEN13606-1:2006 (CEN/TC 251, 2006) and consists of a

Reference Model, an Archetype Interchange

Specification, a Reference Archetypes and Term Lists,

Security Features, and Exchange Models.

The standard defines an architecture for

communicating part or all of the EHRs of a single patient,

making sure that (i) the original clinical meaning intended

by the author of the record is preserved and (ii) the

confidentiality of the data as intended by the author and

the patient is not breached. It does not specify the internal

structure or database design/schema of the EHR.

Part One - The Reference Model.

The Reference Model defines the generic building

blocks of the EHR by representing the global

characteristics of health record components. It shows how

they are aggregated, and how the context information

within the record’s components meets ethical, legal and

provenance requirements. In order to support

communication between various EHR systems, the

standard recommends archetypes, i.e. the use of meta-data

(Frankel, 2006), but they are not mandatory. The

archetypes are “definitions of prescribed combination of

the building-block classes defined in the Reference Model

for particular clinical domains or organizations”

(CEN/TC 251, 2006).

This standard may also be used for communication

between an EHR system and repositories found in clinical

applications, or between middleware components

responsible for accessing or providing EHR data.

Furthermore, the standard may be applied for the

communication of individual patient records (i.e. a single

identifiable subject) to an EHR system or a centralized

EHR data repository. Anonymization or aggregation of

individual patient records within EHR systems are not the

focus of this standard.

The Reference Model consists of four packages –

Extract, Demographics, Support and Primitive. (CEN/TC

251, prEN13606-1:2006).

The Extract package defines the EHR_EXTRACT

class in an EHR structure, which is the root directory of

the Reference Model, and data structures for EHR content.

The Demographics package provides a data set of

specific entities, which are defined only once and then

used within the EHR_EXTRACT by the means of

dedicated class instance identifiers. The whole purpose of

this package is to give a sufficient description of each

entity within an EHR system for human interpretation and

to allow the identification of entities between

EHR_Recepient and its own demographic server. We do

not summarize what Support and Primitives packages

contain because it was very difficult to follow the text

available in the standard.

Part two - Archetype Interchange Specification

The wide-scale sharing of EHR records and their

meaningful analysis across distributed computing sites

requires that the equivalent clinical information is

represented consistently, i.e. when the clinical data

structures are communicated via the Reference Model, it

should have the same semantic meaning. This addresses

the semantic and schematic heterogeneities in the

repositories of EHR systems.

Archetypes introduced in this part have the role of

addressing the challenges of semantic interoperability.

EHR systems should be able to cater for any

professionals, specialities or services within healthcare

systems, despite the fact that healthcare domains are

different, complex and prone to frequent changes. Thus

the Reference Model represents the generic properties of

EHR information, and Archetypes (i.e. meta-data) define

patterns for the specific characteristics of the clinical data

that represent the requirements of each particular

profession, speciality or service. Archetype instances

conform to a formal model available at Archetype Model

prEN13606-2:2005(E)6. Archetypes may also be used

within EHR systems to manage the EHR data, which

belong to a certain data repository. It is assumed that the

original EHR data, if not already archetyped, may be

mapped to a set of archetypes, if needed, when generating

the construct, i.e. EHR_EXTRACT (prEN 13606-2)

Archetype Repositories contain many archetypes that

that have been derived from different sources of EHRs.

Thus despite the fact that different sources of clinical data

inputs, and data types detectable in such inputs, have been

identified, the formats in which these clinical data

structures are represented remain interoperable across a

variety of EHR systems. Consequently, the use of

standardized archetypes ensures an interoperable way of

representing and sharing the specification of archetypes,

in support of keeping EHRs consistent and the semantic

interoperability of shared EHRs.

Communicating Archetypes specifies requirements

for a comprehensive and interoperable archetype

representation, and defines the Object Distributed

Processing (ODP) Information Viewpoint representation

for the Archetype Model. It also gives an optional

archetype interchange format called Archetype Definition

Language (ADL). Archetypes expressed in ADL will be

compatible with HL7 Refined Message Information

Models (R-MIMs) and Common Message Element Types

(CMETs) (Garde, 2007). This part (Communicating

Archetypes) of the standard doesn’t require that EHR

models are archetype prEN 13606-2:2005 compliant.

However it does require that these archetypes are capable

of being mapped to the Archetype Model in order to

support EHR communication and interoperability within

an EHR-sharing community.

Part three - Reference Archetypes and Term Lists

This part specifies data objects for describing rules

for distribution or the sharing of EHRs in whole or in part

by establishing general principles for the interaction of

EHRs with other components and mechanisms within an

EHR application. The standard also establishes ways of

creating information with associated security attributes

(ANSI, 2006).

Part four: Security Requirements and Distribution

rules

This part addresses aspects of data safety and security

in the context of exchanging patient related medical

information. It does this by specifying security

requirements and mechanisms for managing access-rights

to components of an EHR of a patient. It also defines

mechanisms for auditing accesses to an EHR. The

implementation and usability of the security functions of

an EHR system is supported through the provision of

easily implementable but course-grained general access

policies, as well as mechanisms for defining fine-grained,

individual access policies (ANSI, 2005).

Part five: Exchange Models

This part describes a set of models that form the basis

of message-based or service-based communication. Parts

four has already passed approval by CEN but the final

draft is yet to be produced and this part has temporarily

been put on hold.

HL7

The term HL7 is used both: i) as a name for the

organization and ii) as a set of messaging standards

(Version 2.x and Version 3.x). The HL7 organization

focuses on the interface requirements that are needed by

the entire health care organization, when communicating

healthcare data within or outside its healthcare systems.

HL7 standards are the most successful messaging

standards in the healthcare industry. It is a protocol that

consists of standardized grammar and vocabulary. HL7

standards work by assuming that an event in the real

world of healthcare systems creates the need for data to

flow amongst various systems. This will be initiated by a

trigger, which in the HL7 standard is equal to a Trigger

Event (Eichelberg et al., 2005). When an event occurs in

an HL7 compliant system a message is passed to the

requesting application as EDA (Electronic Data

Interchange) by gathering the relevant data from the

applications (Ericson, 2004).

The HL7 standard supports two message protocols:

Version2 and Version3.

HL7 Version2

The most widely used HL7 Version2 protocol is

limited to the exchange of messages between medical

information systems. It was not developed following any

methodology to ensure that all parts of the standard are

developed consistently. HL7 Version2 does not support

interoperability between healthcare applications very

successfully. The main reason for this is the lack of a

precisely defined underlying information model structure,

plus definitions for many data fields are vague and

overloaded with optional data fields (Eichelberg et el,

2005). However, by not defining a detailed information

model, the HL7 version2 standard allows greater

flexibility, which immediately triggers the problem of

interoperability. Applications participating in

communication using HL7 version2 must therefore have

mutual agreements to achieve interoperability.

HL7 Version3

HL7 Version3 is an improvement from the previous

Version2 by being more focused on specific contexts,

terminology, models and conceptual definitions and

relationships. Its underlying information model, called

the Reference Information Model (RIM), is object-

oriented and the proposal for the Clinical Document

Architecture (CDA) for exchanging clinical documents

across healthcare systems uses Extensible Markup

Language (XML) to encode the documents (HL7 2.5,

2000) (InterfaceWare). Thus the CDA defines the

structure and semantics of medical documents that are to

be exchanged and CDA documents use data types

specified in the HL7 RIM.

The CDA consists of 3 levels (HL7 CDA). Each of

them takes the mark up of the previous level and adds

more mark up to compose a clinical document. However,

this does not change the clinical content of the document

(HL7 CDA Release 1.0, 2000)(HL7 Standards).

Level1 consists of a Coded Header and a Body. The

Coded Header defines the semantics of each entry in the

document. The Body contains clinical data in an

unstructured test format or it can consist of nested data

such as paragraphs, lists and tables. Level2 models

observations and instructions within each heading, thus

making it possible to constrain the structure and content of

the document through templates. This increases

interoperability by using agreed templates between

heterogeneous healthcare systems. Level3 provides

completely structured documents where each element of

the document is adequately coded for machine processing.

The CDA HL7 is not strictly an EHR standard, but

forms its sub-component, that has already been

harmonized with the equivalent structure in CEN13606

and openEHR (HL7 EHR 2004). Thus HL7 and CEN co-

operate and the current areas of harmonization include

data types from (a) HL7 Templates (b) HL7 CDA, (c)

CEN13606 Reference Models, and (d) CEN/openEHR

archetypes (Schloeffel P., 2004).

Digital Imaging and Communications in Medicine

(DICOM)

DICOM is the de facto standard for medical image

communication. It uses binary encoding with hierarchical

lists of data elements identified by numerical tags and a

complex DICOM-specific application level network

protocol (DICOM, 2007). There are two DICOM-based

EHR standards available:

1. Web Access to DICOM Persistent Objects (WADO)

2. DICOM Structured Reporting.

WADO is a cooperative standard between DICOM

(DICOM Supplement 85 2004) and ISO (ISO 17432

2004). This standard was created to enable and maintain

international standards for the communication of

biomedical diagnostic and therapeutic information in

disciplines that use digital images and associated data.

WADO defines a Web based service that can be used to

retrieve persistent objects via HTTP or HTTPS from a

Web server.

The Web client has to specify the DICOM object that

has to be retrieved. This will be done through unique

identifiers available at the instance level of the DICOM

information model. The web client may also request a

specific data type to be sent via the web server. The web

server will convert the existing DICOM object into a

readily presentable format to be sent. The DICOM object

can also be made anonymous before it is sent. This is a

very useful feature for teaching purposes and clinical

studies. WADO doesn’t support query mechanisms.

A WADO server will return a DICOM Structured

Reporting Document to the client in the HTML format if

requested. If not, the format will be dependent on the

current implementation of the server. WADO uses a

simple approach for accessing particular DICOM objects

without requiring the client to be DICOM compatible. It

is very easy to implement applications supporting WADO

as they can be implemented by using readily available

components such as Web browsers, servers and DICOM

viewers (Eichelberg et el, 2005).

DICOM Structured Reporting is a general model for

encoding medical reports in a structured way. This is

done using DICOM’s tag-based format. It allows for

current DICOM infrastructure network services to be used

to archive and to communicate, and to digitally sign

structured reports without making significant changes to

the existing systems.

To improve interoperability in practice, DICOM

Structured Reporting specifies a document structure along

with its class definitions and constraints for different

medical applications. It defines templates that need to be

used for this purpose. The collection of the standard

templates, context groups and codes is called the DICOM

Content Mapping Resource.

It is important to note that DICOM does not specify

how a Structured Reporting Document is rendered by an

application, as this is regarded to be out of the scope of

the standard (Eichelberg et el, 2005).

HARMONIZATION BETWEEN STANDARDS

Harmonization between standards and in some cases

their convergence, is very important for the future

interoperability of health information systems within and

between countries. As we have already mentioned in the

previous section, current areas of harmonization include

data types: the HL7CDA and CEN 13606 Reference

Models and CEN/openEHR archetypes with HL7

Templates. The OpenEHR Reference Model uses the

CEN13606 Reference Model, which in turn is used in

HL7CDA. So, it is quite visible that some harmonization

has already taking place. There are also sporadic reports

across EU countries that highlight their individual

attempts to comply with CEN initiatives, thus

recommending harmonization locally (SIST). A detailed

rationale for the harmonization of EHR standards is

available in the white paper about the HL7 EHR System

Functional Model (HL7, 2004).

CONCLUSION

It was very difficult to prepare an overview of EHR

interoperability standards that could have provided better

reading than this paper offers. It was also extremely

challenging to extract information that can be called ‘an

overview of standards’ from such a variety of sources,

where almost none of them are written for wider

healthcare or computing research communities. There are

a few published papers, which have helped us to interpret

the texts found in published standards correctly and which

have been crucial to understanding a variety of issues

covered in these standards. Thus any organizations,

healthcare institutions or software developers willing to

include these standards in order to achieve the

interoperability of EHR systems, will face the same

difficulties as we have had in collecting relevant material

and understanding how to apply it. One of the solutions is

to employ professionals who can interpret the standards so

that they can be understood by software developers and

healthcare professionals.

We do not wish to criticize the organizations which

have produced these standards but we do think that their

contribution towards solving the interoperability of EHR

systems is enormous. However, it has become quite

obvious that there is no consistency or coherence in what

these standards cover: some of them are concerned with

the content and structure of EHRs and others deal with

access services to such records. It is easy to claim that the

main purpose of EHR standards is to facilitate

improvements in the five areas of interoperability,

safety/security, quality/reliability, efficiency

/effectiveness, and communication. However, when they

advocate how to achieve comprehensive EHR in practice,

they advise to join together multiple clinical applications

and their specially tailored databases (HL7 EHR, 2004)

(ISO TC 215, Draft 2002). This means that EHRs from

diverse systems must be capable of being mapped to and

from single comprehensive representation. Furthermore,

this common representation is expected to be generic

enough to represent any kind of health record which might

be a partial or complete EHR being communicated.

We know how difficult is to address the semantic and

schematic heterogeneities in software applications that

rely on data stored in structured and semi-structured

repositories and databases which are often heterogeneous.

Thus it is a surprise that none of these standards took into

account experiences from software engineering and

database communities, which have tried to alleviate the

interoperability problems since the early 1990s. For

example, requiring uniformity in heterogeneous database

systems diminishes their individual autonomy and

discourages their heterogeneities, which is exactly what

we have to have in modern pervasive and ubiquitous

healthcare environments, served by EHRs.

Having said that, HL7 version3 can be considered to

be the foundation of future integrated health care

environments. By combining HL7 version3 and the CEN

ENV 13606 standard, organizations can get a complete

standard solution for the introduction of Information

Communication Technology into the health care domain.

There is definitely a space for integration between various

healthcare systems and applications, which can be

achieved by providing a transparent exchange of health

care information. From this point of view, all the

standards meet this specification. However, from the

software engineering perspective, interoperation of such a

system is more complex than standards could define.

It is not possible to favor one standard over the

other. They are all quite similar in what they cover,

however at different stages of their evolution. There will

always be organizations that will be using incompatible

EHR standards. The only solution will be when full

semantic interoperability is reached.

REFERENCES

Akram A., Juric R., Ranganathan M., Slevin L.

“

Databases for facilitating data sharing in The UK NHS”,

accepted for Integrated Design and Process Technology, 2007.

ANSI, 2005, “Health Informatics - Electronic health

record communication - Part 4: Security requirements and

distribution rules”, version prEN 13606-4:2005, ANSI,

USA.

ANSI, 2006, “Health informatics - Electronic health

record communication - Part 3: Reference archetypes and

term lists”, version prEN 13606-3:2006, ANSI, USA

Beale T., 2002, “Archetypes: Constraint-based

Domain Models for Future-proof Information Systems”,

Deep Thought Informatics PTY, LTD Mooloolah, Qld,

Australia.

Bossen C., 2006, “Representations at Work: A

National Standard for Electronic Health Records”.

Proceedings of the 2006 20th anniversary conference on

Computer supported cooperative work, Banff, Alberta,

Canada, pp 69 – 78.

Bott O.J., 2004, “The Electronic Health Record –

Standardisation and Implementation”, 2

OpenECG

Workshop, Berlin, Institute for Medical Informatics,

www.ifmi.de.

Cabinet Office, 2005, e-GIF e-Government

Interoperability Framework, Technical Standards

Catalogue Version 6.2

http://www.govtalk.gov.uk/schemasstandards/egif_docum

ent.asp?docnum=957,

CEN/TC 251, prEn 13606-1:2006, “Health

Informatics – Electronic Health Record communication,

Part 1: Reference Model, European Standard”.

http://www.chime.ucl.ac.uk/resources/CEN/EN13606-

1/N06-02_prEN13606- 1_20060209.pdf

CEN/TC 251, ENV 13606-1, 1999, “Electronic

healthcare record communication – Part1: Extended

architecture”, Committee European Normalisation, Health

Informatics Technical Committee.

CEN/TC 251, prEN 13606-2, 2005, “Health

Informatics - Electronic health record communication —

Part 2: Archetypes”.

CEN/TC 251, prEN 12967-3, 2006, “Health

informatics — Service architecture — Part 3:

Computational viewpoint”, www.centc251.org.

CEN/TC 251, Working Group IV Technology for

Interoperability, www.tc251wgiv.nhs.uk

DICOM, 2004, “Part1: Introduction and Overview”,

PS3.1. National Electrical Manufacturers Association,

Virginia, USA

DICOM, 2007, “Part 2: Conformance”, PS 3.2 – 2007,

National Electrical Manufacturers Association,

Virginia, USA

DICOM, Supplement 85, 2004. Web access to

DICOM persistent objects (WADO). Joint DICOM

Standards Committee, ISO TC215 Ad Hoc Working

Group on WADO.

Dumay A.C, Ferkins G., 2002, “Towards developing a

coherent healthcare information infrastructure”, Studies in

Health ogynformatics, Vol93, pp 1-7.

Eichelberg M., Aden T., Riesmeier J., Dogac A.,

Laleci G., 2005, “A Survey and Analysis of Electronic

Healthcare Record Standards”, ACM Computing surveys,

Vol37(4), pp277-315.

Ericson N, 2004, “Health Care Agent”,

http://www.ericsson.com/hr/products/e-

health/HC_Agent_R1B.pdf

Frankel H., 2006, “Archetypes: Update”, 10

conference HL7, Australia.

Garde S, Knaup P, Hovenga EJS, Heard S, 2007,

“Towards Semantic Interoperability for Electronic Health

Records: Domain Knowledge Governance for openEHR

Archetypes”, Methods of Information in Medicine,

Vol.46,(1).

Giuse D.A., Kuhn K.A., 2003, “Health information

systems challenges: the Heidelberg conference and the

future”, International Journal of Medical Informatics,

Vol.69 (3), pp 105-114.

Gross G., 2005, “Lack of standards hinders electronic

health records. Interoperability concerns loom large.”,

IDGNS. www.infoworld.com.

Health Level7. http://www.hl7.org/

HL7, 2004, EHR System Functional Model: “A Major

Development Towards Consensus on Electronic Health

Record System Functionality”, A White Paper. Health

Level Seven, Inc.

HL7, 2005, Application Protocol for Electronic Data

Exchange in Healthcare Environments, Version2.5.

HL7, 2000, “The HL7 Version 3 Standard: Clinical

Data Architecture”, CDA Release 1.0.

Hsyien-Chi W., Yuh-Shan H., Wen-Shan J., Hsien-

Chang L., Yi-Hsin E.H., 2007, “Scientific production of

electronic health record research, 1991-2005”, Computer

Methods and Programs in Biomedicine, Vol.86, pp191-

196.

InterfaceWare, www.interfaceware.com/hl7.html

ISO, TC 215, 2002, EHR ad hoc Task Group,

“Standards requirements for the EHR”, Draft 1.2.

http://secure.cihi.ca/cihiweb/en/downloads/infostand_ihis

d_isowg1_e_EHR_TaskGroup.pdf

ISO, http://www.iso.org/iso/en/aboutiso/introduction

/index.html

ISO TC215, ISO/TR 20514, 2005, “Health Informatics

– Electronic health record – Definition, scope, and

context”, ISO, Geneva, Switzerland.

ISO 17432, 2004, “Health Informatics –Messages and

Communication – Web Access to DICOM Persistent

Objects”, ISO, Geneva, Switzerland.

ISO/TR 18307, 2001, “Health Informatics –

Interoperability and Compatibility in messaging and

communication standards – Key characteristics.”, ISO,

Geneva, Switzerland.

Kalra D., Ingram D, “Electronic Health Records”,

Centre for Health Informatics and Multiprofessional

Education, UCL, London, UK

Kalra D., 2006, “Electronic Health Record Standards”,

IMIA Yearbook of Medical Informatics, pp 136-144.

Kuhn K.A., Giuse D.A., 2001, “From hospital

information systems to health information systems,

problems, challenges perspectives, Methods of

Information in Medicine. Vol. 40(4), pp275-287.

Medisell, 2002, “Interoperability with EPR systems:

Standards, Technologies & Trends”, Information Society

Technologies.

NEMA, National Electrical Manufacturers

Association. www.nema.org.

NPfIT, 2004, National Program for IT, Care Record

Development Board, Ethics Advisory Group, Ethical

Principles

http://www.connectingforhealth.nhs.uk/crdb/boardpapers/

docs/crdb_ethical_prin.pdf

openEHR, EHR Standards,

www.openehr.org/standards/t_iso.htm.

The OpenEHR Foundation,

http://www.openehr.org/about_openehr/t_home_aims.htm

Schloeffel P., 2004, “Current EHR Developments: An

Australian and International Perspective.” Health

Informatics New Zealand Conference (HINZ2004),

Wellington, New Zealand.

Schloeffel P., Jeselon P., 2002, “Standards

requirements for the EHR”, Discussion Paper, Draft 1.2.

SIST, Slovenian Institute for Standardization,

Ljubljana, Slovenia, www.sist.si.

Slevin, L., Dagdeviren, H., Juric, R., and Akram A.,

(2005) "Supporting Quality Indicators in the UK National

Health Service", In Proceedings of the 27

International

Conference on Information Technology Interfaces, ITI

2005 (Cavtat, Dubrovnik, Croatia, 20-23 June 2005).

Ueckert F., Goerz M., Ataian M., Tessmann S.,

Prokosch H.U., 2003, “Empowerment of patients and

communication with health care professionals through an

electronic health record, International Journal of Medical

Informatics, Vol.70 (2-3), pp 99-108.

Valdes I., Kibbe D.C., Tolleseon G., Kunik M.E.,

Petersen L.A., 2004, “Barriers to proliferation of

electronic medical records”, Informatics in Primary Care,

Vol.12, No 1, pp3-9.

Van Ginneken A.M., 2002, “The computerized patient

record balancing effort and benefit.”, International

Journal of Medical Informatics, Vol.65 (2002),pp 97-119.