See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/6688502

The International Classification of Primary Care (ICPC-2): an essential tool in

the EPR of the GP

Article  in  Studies in Health Technology and Informatics · February 2006

DOI: 10.3233/978-1-58603-647-8-809 · Source: PubMed

CITATIONS READS

65 4,118

4 authors, including:

Diego Schrans Jan De Maeseneer

Ghent University Ghent University
13 PUBLICATIONS   122 CITATIONS    370 PUBLICATIONS   8,055 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

people centred care View project

Climate change - Migration - Health(care) nexus View project

All content following this page was uploaded by Jan De Maeseneer on 04 June 2014.

The user has requested enhancement of the downloaded file.

The International Classification of Primary
Care (ICPC-2): an essential tool in the EPR
of the GP.
Marc Verbeke M.D.1, Diëgo Schrans M.D., Sven Deroose M.D., Jan De Maeseneer
M.D. Ph.D
Department of General Practice and Primary Health Care,
Ghent University, Belgium

Abstract. The International Classification of Primary Care (ICPC) has become a

standard all over the world. It became a standard tool to classify the important
elements in the Electronic Patient Record (EPR) of the GP: reasons for encounter
(RFE) reflecting the patient’s view, process of care (decision, action, intervention
or plans) reflecting the care process, and the assessment (diagnosis or health issue)
reflecting the doctor’s view. ICPC-2 is fully compatible with structuring data in
the episode of care model and it’s reflecting the essential elements of each
patient/provider encounter. To implement ICPC-2 in the EPR a Thesaurus has
been developed in Belgium with double encoded clinical labels. The
implementation is now mandatory for labeled EPR systems in Belgium. The use of
ICPC 2 may improve the accessibility and use of on-line Expert systems and
Guidelines.
Keywords: Medical Informatics, ICPC, primary care, thesaurus, medical record.

1. Introduction

The International Classification of Primary Care (ICPC), developed by the ICPC

Working Party, broke new ground in the world of classification when it was published
for the first time in 1987 by WONCA [6], the World Organization of National
Colleges, Academies, and Academic Associations of General Practitioners/Family
Physicians, now known more briefly as the World Organization of Family Doctors
(Wonca). For the first time health care providers could classify, using a single
classification, three important elements of the health care encounter: reasons for
encounter (RFE), diagnoses or problems, and process of care. Problem orientation of
the medical record and linkage of encounters over time enables classification of the
episode from the beginning with an RFE to its conclusion with a more defined
problem, diagnosis, or disease.

1
Corresponding Author: M.Verbeke UZ.1k3, De Pintelaan 185, B9000 Gent, Belgium.
Marc.verbeke@Ugent.be.
2. The choice of the International Classification of Primary Care (ICPC)

A classification relevant to primary care is necessary. It should be logically organized

to:

- reflect the concerns of patients (not the immediate assumptions or beliefs of

their doctors)
- characterize episodes of care
- be easy to routinely use
- be inexpensive
- be relevant to any primary care setting
- be expandable
- be connected to existing coding and other classification approaches
- be readily deployed in the evolving electronic health record.

ICPC with its linkage to ICD-10 which is sanctioned by World Health Organization
and included in the US National Library of Medicine’s Unified Medical Language
System (UMLS) is the only existing classification scheme that meets these standards
[9], [10].Using ICPC with it’s ordering principle in a defined domain and based on
the high prevalence of common diagnoses in family practice, gives much more power
and possibilities in care management, than could ever be reached by using only
terminologies or vocabularies as SNOMED or Read codes. Another reason to prefer
ICPC is the international recognition and the inclusion in the WHO-FIC (WHO Family
of International Classifications). ICPC acts as the link between other classifications
(ICD, ICF, ICHI, ATC) in the FIC.

2.1. History and future.

The first version of ICPC published in 1987 is referred to as ICPC-1 [6]. In 1998
Wonca published a revised version: ICPC-2 [7] with inclusion and exclusion criteria
attached to the classification rubrics, and a mapping to ICD-10. ICPC-2e is an
electronic version from year 2000 of the revised and corrected chapters 10 and 11 of
the ICPC-2 book [13]. In 2003 WHO recognized ICPC-2 as a WHO related
classification for the recording of data in primary care.
A new version numbering system was introduced by the WICC Update group in
November 2005. The latest version of ICPC-2e will always be available at the web site:
http://www.kith.no/templates/kith_WebPage____1062.aspx [2]. Information about
ICPC history, background, member countries etc. can be found on the WICC web
pages [1].
ICPC-3 will be developed in the coming years: nothing has been decided yet but
there are some accepted points of discussions:
- ICPC structure will probably be maintained due to
relationships with ICD within WHO-FIC
- including patient preferences
- mapping to ICF and/or including new rubrics for this?
- process codes to be revised and maybe including
objective findings and investigation
- revising prevention rubrics
 - mapping to other classifications?

2.2. Structure of ICPC

ICPC is based on a simple bi-axial structure: 17 chapters based on body systems on one
axis, each with an alpha code, and seven identical components with rubrics bearing a two-
digit numeric code as the second axis.
Component 1 provides rubrics for symptoms and complaints. Rubrics in this
component can be used to describe presenting symptoms, and are valuable for describing
the problem under management when the condition is as yet ill defined. Component 7 is the
diagnosis/disease component in each chapter. This component will be the one most often
used when you have sufficient information to arrive at a diagnosis in the medical record.
Within this component 7 are five subgroups, which are not numerically uniform across
chapters: infectious diseases, neoplasms, injuries, congenital anomalies, and other diseases.
Components 1 and 7 in ICPC function independently in each chapter and either can be used
to code patient RFEs, presenting symptoms, and diagnoses or problems managed.
Components 2-6 (process-codes) are common throughout all chapters, each rubric being
equally applied to any body system.
The structure of ICPC represents a move away from the combined anatomical and
etiology based structure of ICD. For example, where ICD includes a separate chapter
for neoplasm’s, one for infections and infestations, and another for injuries, such
problems are distributed among chapters in ICPC, depending on the body system to
which they belong.

3. Belgian Bilingual Thesaurus and clinical labels

Primary care physicians have an immediate need for a simple and honest way to
routinely record and retrieve data reflecting their perspective. The primary care
perspective must be incorporated into the nation’s data standards and electronic health
records. Clinical research and a fully integrated health information system cannot be
sustained without practical, easily used primary care data standards.
Because of the mandatory implementation of ICPC/ICD classification systems
within the labeling procedure in 2006 for Belgian software systems in primary care, a
user-friendly instrument to encode data had to be created in order to make
classification systems acceptable and useful. Implementing only a classification
without search instruments would never be accepted by family physicians who don’t
know anything about coding systems. Coding data in the ‘background’ of the EPR and
showing only the clinical label as the doctor wants it, is the best way to make it
acceptable. For this reason, under the authority of the federal government, a bilingual
Belgian Thesaurus has been developed in use of GP’s reporting and encoding the data
in the EPR. A fruitful collaboration of two university departments of Primary Care
(Ghent University Ugent for the Dutch part and Brussels University ULB for the
French part of the Thesaurus) has been installed for this purpose. Implementation of the
Thesaurus with double encoding system is mandatory for the labeling of the EPR
system by the government in 2006 [5].
The Belgian labeling system is a way to guarantee quality, inter-operability,
communication possibilities and standards in the EPR systems in combination with a
free market of software systems in Health Care. The doctor gets financial incentives for
using a software system that has been labeled by the government.

3.1. Position of a Thesaurus in reporting and organizing data in the EPR

A Thesaurus creates a bridge between the words (or in the EPR: parts of words) in the
GP’s thoughts in describing symptoms, complaints, assessment or other data to be
reported in the patient record and the scientific correct way of describing clinical
concepts, used in organizing the record-system with all it’s possibilities of
communication, research and other management issues.

Translating steps:
- doctor’s vocabulary linked to
- terminology of clinical concepts used in reporting data
in EPR and linked to
- clusters of concepts linked to
- classification systems:
o primary care ICPC
o ICD if more granularity is needed
- Organization of EPR (prevention, risk management,
expert systems etc.) based on these classifications in use.
- Classifications linked to guidelines, research,
administration, insurance systems and communication
with other health care providers

3.2. Development of a Thesaurus

The starting point of this Belgian Thesaurus was a Thesaurus created in

Amsterdam by the department of Prof. H. Lamberts (UvA), in collaboration with Ghent
University, Department of General Practice and Primary Health Care [11]. But this
Thesaurus has been refused as such by developers and users especially because of the
lack of clinical labels useful as text in the EPR. Search terms and ICD/ICPC labels (the
content of the Amsterdam Thesaurus) are not always the preferred text in a medical
record. The Belgian GP needs also a perfect bilingual system by which all information
(codes and labels) became completely interexchangeable from one language to another
(French/Dutch) and from one software system to another.
For this reason each collection of search terms (doctor’s vocabulary) has been
linked to a specific ‘clinical label’ (terminology of clinical concepts) useful as
terminology in the EPR and to the appropriate pair of classification codes (clusters of
concepts): ICPC-2-ICD10. Each clinical label is perfectly bilingual French/Dutch. The
search terms are specific but different for each language. The classification mapping is
according the rules of the Wonca International Classification Committee (WICC) [8].
In order to avoid as much as possible “double or identical labels” that could be
linked to different pairs of codes, we have rewritten the labels in a way each
description, each label has been formulated according the content of the mapped ICD
/ICPC codes. This system improves the accuracy of the chosen label and mapped codes.
In ICPC chapter X and Y (female and male genital systems) a lot of labels are identical,
but the EPR can know which one has to be taken according patients gender.
According to ICPC philosophy we prefer to encode diagnosis as ‘manifestation’
instead of ‘etiology’. Example: ‘tuberculous prostatitis’ has been mapped to ICPC/ICD
Y99/N51.0 (Genital disease male other/Disorders of prostate in diseases classified
elsewhere) but not to A70/A18.1 (Tuberculosis/Tuberculosis of genitourinary system).
This point of the Thesaurus has to be evaluated by the users to know if both viewpoints
(etiology and manifestation) have to be implemented or not. The disadvantage to keep
both would be two identical ‘clinical labels’ expressing the same concept but with
different codes. Within the context of clinical research, this could be an important issue.

3.3. Results [4]

About 49.000 different clinical labels have been created. Most of them are unique and
compatible with the content of the appropriate pair of ICPC/ICD codes. By using this
system the GP has a vocabulary and terminology system at his disposition, which
encodes the reported data automatically without the GP necessarily knowing the codes.
This encoded information is useful to organize patient record systems for prevention
(inclusion and exclusion criteria to a specific preventive program), risk calculation, bi-
directional communication with other health care providers and even expert systems as
access to guidelines, working flow charts, assistance in prescriptions etc.
Besides this creation of a bilingual and double encoded Thesaurus with added
clinical labels, expert systems have been indexed with ICPC codes to make it ‘on-line’
accessible by using the assessment (the ‘A’ of SOAP) codes from the EPR. This makes
guidelines accessible with a language independent system, avoiding typewriting errors
and incorrect search terms. The Belgian GP can use the French or Dutch search terms
from the Thesaurus to search in English guidelines, in this case: the ‘EBM Guidelines’
from Duodecim in Finland [3]. Even process codes (in ‘O’ or ‘P’ of SOAP) can be the
trigger of access to the guidelines. The key for this link is the used ICPC code, but
maybe in the future also the mapped ICD code could be used. A correct granularity
offered by ICPC and the mapping to ICD is an essential tool to organize this, and can
hardly be done, if not impossible by using only terminologies, vocabularies or ICD.
Also procedures are encoded with a list of about 1500 different most common
procedures done or ordered by the GP. In ICPC component 2-6 all codes are chapter-
independent. For this reason the same rule has been respected and all of the 1500
procedures has been added chapter independent as extensions of the ICPC process
codes *30 to *69. The encoded procedures are very helpful in organizing the process of
taking care by the GP and in Belgium the GP has certainty that all procedures
reimbursed to the patient and reported mandatory in his EPR, are within that list. This
makes the list of procedures quite adapted to the Belgian situation, but could very
easily be changed to fit for other countries where GP’s are doing different things.

3.4. Future developments.

A thesaurus useful in daily practice will never be finished: medical concepts are
changing, new concepts are developing, other information can become more important,
language is dynamic and GP’s are dynamic, search terms will be missing and errors
will be detected, classifications will be updated, other classifications could be added
 (i.e. ICF: International classification of functioning, disability and health) [12] and so
on, so the Thesaurus will always be changing.
Other classification systems will be added in communication, administration and
organization of health care. The Thesaurus is only a part of the tool the GP needs
within the care process, but it’s an important tool in the organization of the EPR and it
improves the possibility to a better quality of care. The better quality of encoded data
means also an improvement of epidemiologic data and scientific studies in primary
care in Belgium.

3.5. Conclusion

A Thesaurus is a good instrument to give GP’s the freedom of reporting in the EPR the
data in their own language and words without loosing all the opportunities and help in
the management of taking care offered by language independent classification and
coding systems. Even GP’s who are ‘allergic’ to codes and classifications are pleased
by the extra value offered by using a Thesaurus.

Acknowledgment

The development of the Belgian Thesaurus has been made possible by the grant of the
Belgian Federal Administration of public health.

References

[1] http://www.globalfamilydoctor.com/wicc/icpcstory.html
[2] http://www.kith.no/templates/kith_WebPage____1062.aspx
[3] http://www.ebm-guidelines.com (last accessed Sept. 30, 2005).
[4] www.health-telematics.be/thesaurus.
[5]
https://portal.health.fgov.be/portal/page?_pageid=56,4280434&_dad=portal&_schema=PORTAL&_me
nu=menu_5_6_1
[6] Lamberts H, Wood M, eds. ICPC. International Classification of Primary Care. Oxford: Oxford
University Press, 1987.
[7] ICPC-2. International Classification of Primary Care. Second edition. Oxford: Oxford University Press,
1998.
[8] International Classification of Primary Care ICPC-2-R, Revised second edition, WONCA International
Classification Committee, Oxford University Press 2005; ISBN 978-019-856857-5
[9] Banff Declaration calls for ICPC in the United States. Wonca News, April 2004. Okkes IM.
[10] The Banff declaration and the information needs of Primary Care and Family Medicine in the USA.
http://annalsfm.highwire.org/cgi/qa-display/short/annalsfm_el;48#175, 16 Dec 2003
[11] Becker HW, Van Boven C, Oskam SK, Hirs WM, Lamberts H, Verbeke M, et al, eds. ICPC2-ICD10
Thesaurus. Version May 2003. Amsterdam, Den Haag etc: Academic Medical Center/University of
Amsterdam, Ministry of Health, Dutch College of General Practitioners, Dutch Institute for Public
Health and Environment RIVM, Free University of Brussels, University of Ghent, 2003.
[12] ICF publication by WHO: International classification of functioning, disability and health, ICF. Geneva
2002
[13] Okkes IM, Jamoulle M, Lamberts H, Bentzen N. ICPC-2-E. The electronic version of ICPC-2.
Differences with the printed version and the consequences. Fam Pract 2000; 17: 101-6.

View publication stats