WHO - FAMILY OF INTERNATIONAL CLASSIFICATIONS NETWORK ANNUAL MEETING 2013

ICD-11 and the “Factors influencing health status and 12 – 18 October 2013
contact with health services”: a test of integration for the Beijing, China
Family of International Classifications
C301

Gongolo F.1, Bang S. 2, 3, Sykes C. 4, 5

1Central Health Directorate of Friuli Venezia Giulia Region – Italian WHO-FIC Collaborating Centre; 2 Statens
Serum Institut, National e-Health Authority, Denmark, 3 Collaborating Centre for the WHO-FIC in Nordic
Countries, 4 World Confederation for Physical Therapy (WCPT) 5. Faculty of Health Sciences, University of Sydney

Abstract This poster represents a proposal for the restructuring of ICD-10 Chapter XXI within the ICD-11 revision process .

WHO-FIC
Introduction ICD-10 code and title
classification
Contact with and
ICD-10 Chapter XXI is used to record health ICD-11 post-
Z20.5 exposure to viral
related circumstances that are not a hepatitis
coordination
disease. These categories, coded with Z Contact with and
codes, contain a mix of concepts, including exposure to human ICD-11 post-
reasons for encounter, risk factors and Z20.6
immunodeficiency coordination
interventions. The current scenario of the virus [HIV]
revision of ICD presents an opportunity to Contact with and
review the chapter and propose new ways exposure to ICD-11 post-
Z20.7
of organising the content. The Topic pediculosis, acariasis coordination
Advisory Group on functioning (fTAG) has and other infestations
the mandate of suggesting a new structure Tab. 2 – Examples of Z codes as ICD-11 post-
of the Z codes chapter starting from the coordination dimensions
alignment of the revised ICD and ICF but
also taking into account all the possibilities
given by the joint uses of the WHO-FIC.

Methods & Materials Fig. 2 – a screen shop of Z codes in iCAT

After a preliminary meeting of fTAG co-
chairs with WHO a Z codes working group
was formed. Background materials and
briefing notes were made available and a
work plan drafted. Current debate on Z
codes and suggestions for change were
collected through a search of scientific
literature and through the commenting
features of the WHO ICD-11 beta browser.
The ICD-10 chapter XXI (current ICD-11
Chapter 23, Fig.1), exported in spread sheet
format from the revision collaborative
platform (Collaborative authoring tool, iCAT,
Fig. 2), was taken as starting point for
redrafting the chapter. The different blocks
of chapter XXI were reviewed in
teleconferences, highlighting the relevance
of the single classification entities in terms
of their relevance as post-coordination
categories of ICD-11, as contextual factors
of the International Classification of
Functioning Disability and Health (ICF), or
as categories of the International
Classification of Health Interventions (ICHI,
now under development).
(June 2013)
Results
Of the 801 ICD-10 chapter XXI categories
considered in the analysis 158 ICD-10
categories relate to ICF contextual factors.
In this regard the development and
integration of a personal factors
Fig. 3 – Overall possible reassignments of
classification is encouraged. A large group of former ICD-10 Chapter XXI Codes
categories (367) could be represented in an
interventions classification. Classifications of
devices and assistive technology are Conclusions
important extensions to these categories.
The possibility offered by ICD-11 to post- Revising the structure of ICD-10 chapter
coordinate dimensions such as “history of” XXI in the context of development of ICD-11
would make redundant another 105 offers the possibility of testing the actual
categories (see examples below in Tab. 1 integration of the Family of International
and 2). For 171 categories alternative Classifications in terms to effectively
possibilities for ordering the concepts represent, beyond the disease, all
remain to be debated. Overall results are dimensions of health. In order to achieve
shown in Fig. 3 this task the fTAG highly values the
involvement, through the collaborating
WHO-FIC
ICD-10 code and title centres, of experts in all WHO reference
classification
classifications.
ICHI/interventions
Z51.0 Radiotherapy session related
Acknowledgements

Z51.2
Chemotherapy ICHI/interventions Authors are members of the fTAG and in
session for neoplasm related their work took advantage of preparatory
Blood transfusion materials made available by the Group for
(without reported ICHI/interventions this purpose.
Z51.3
diagnosis) related

References
Accentuation of PERSONAL 1. Kennedy C. Overview of ICD-10 Version
Z73.1
personality traits FACTORS
2010 Z Codes V1 3.21.2013
Lack of relaxation and PERSONAL
Z73.2 2. iCAT
leisure FACTORS
PERSONAL http://icat.stanford.edu/
Z73.3 Stress
FACTORS
3. ICD-11 Beta browser
Inadequate social PERSONAL
Z73.4
skills FACTORS http://apps.who.int/classifications/icd11/browse
PERSONAL /f/en
Z73.5 Social role conflict
FACTORS
DIGITAL, MOBILE, NOW!
Fig. 1 – The Morbidity linearization of ICD-11 Presence of artificial ICHI/interventions
Z97.0
Chapter 23, former ICD Chapter XXI (June eye related+ ISO9999
2013). Tab. 1 – Examples of Z codes and relevant Scan this to get a digital version
WHO classifications
 WHO - FAMILY OF INTERNATIONAL CLASSIFICATIONS NETWORK ANNUAL MEETING 2013

12 – 18 October 2013
A Semantic Web-base Hybrid Solution for Beijing, China

Lexical Term Standardization of ICD-11 C302

Guoqian Jiang, Harold Solbrig, Christopher Chute

Mayo Clinic College of Medicine, Rochester, MN
Abstract
As part of work plan of the World Health Organization (WHO), the lexical terms used in ICD-11 labels or properties of
entities should follow a standard and homogeneous approach. The objective of the study is to develop and evaluate a
Semantic Web-based hybrid solution to identify term heterogeneity in ICD titles.
Title
Verdana 36 Bold
INTRODUCTION NLM Sub-Term Mapping Tools (STMT)
The beta phase of the 11th revision of International Services to find all sub-
Services to find all
Classification of Diseases (ICD-11) started in May 2012. Lexical Sub-term permutations of
terms for an entity
Semantic Web synonymous sub-terms
As part of work plan of the World Health Organization Services
(WHO), the lexical terms used in ICD-11 labels or WHO ICD-11 Content Services
properties of entities should follow a standard and (Entity titles, definitions and values)

homogeneous approach. For example, the terms

“cardiac”, “renal” should always be used in ICD titles,
definitions and values instead of the terms “heart”,
“kidney”. The objective of the study is to develop and Lexical Sub-term
Data Viewer
Lexical Sub-term Data Repository
evaluate a hybrid solution to identify term heterogeneity
Title (Semantic Web RDF Triple Store)

in ICD titles.
Fig. 1 System Architecture
METHODS & MATERIALS
Fig.1 shows the system architecture. We utilized a
lexical toolset known as the Sub-Term Mapping Tools
(STMT) developed at National Library of Medicine (NLM).
Two main features of the tool were used: 1) to find all
sub-terms for an entity; 2) to find all permutations of
synonymous sub-term substitutions. We developed a
Semantic Web-based wrapper service that links WHO
ICD-11 content services with local STMT lexical sub-term
services. Specifically, the wrapper service takes an ICD
entity URI as the input (which retrieves the title of the
ICD entity), and then renders the sub-terms of an entity
and the synonyms of each sub-term in a Semantic Web
Fig. 2 A Sub-term Pair Example (heart/cardiac)
Resource Description Framework (RDF) format using the
W3C standard Simple Knowledge Organization System
(SKOS) signatures (Fig. 2). Using the wrapper service,
we harvested the sub-terms and their synonyms in RDF
triples for all foundation entities (n= 29,445) and loaded
them into an open source RDF triple store known as
4store. We enabled a SPARQL endpoint that provides
standard SPARQL query services against the sub-term
dataset. We analyzed the dataset using a set of SPARQL
queries (Fig. 3).

RESULTS Fig. 3 A SPARQL Query for Sub-term Pair Extraction

As a pilot study, we retrieved the sub-term pairs, in Conclusions
which the preferred label of a sub-term appears to be
the synonym of the other sub-term, and identified 4,927
distinct sub-term pairs. We manually reviewed a small
subset of the sub-term pairs and concluded that they
reflect reasonably well the term heterogeneity in ICD
titles. For example, for the sub-term “bleeding”, we
identified its synonymous sub-terms “haemorrhages”,
“haemorrhage”, “haemorrhagic”, “hemorrhage”,
“hemorrhagic”, “blood loss”, “ruptured”, “rupture”, “spot”,
and “spotted”. We plan to work with WHO to conduct a
comprehensive review for the set of sub-term pairs and
their frequency distribution in ICD titles (Fig. 4).

CONCLUSION
In summary, we developed a hybrid approach that
combines an NLP-based lexical tool with a semantic web-
Fig. 4 A Sub-term Data Viewer (with frequency distribution)
based approach, which would provide an effective and
scalable solution for lexical term standardization of ICD-
11.
WHO Academic Collaborating Centre
at Mayo Clinic (under designation)
 WHO - FAMILY OF INTERNATIONAL CLASSIFICATIONS NETWORK ANNUAL MEETING 2013
Clinical Dictionary for
iSalut
Authors: Rius A, Graupera A, Nozal M, Rovira M, Gallego C,
García F, Cornet J
Abstract The project of Clinical Dictionary for iSalut has the objective to standardize the vocabulary used in the Catalan
Health System to allow semantic interoperability between systems of its different healthcare providers. The base of the
dictionary is the international terminology SNOMED CT and its domains are work as subsets of this standard. The dictionary
covers diverse areas as immunizations, allergies, groups of professionals or referral procedures, and keeps growing.
Introduction
The Clinical Dictionary project is a part of iSalut, a
transversal and strategic program that aims to transform the
healthcare assistance model in Catalonia (Spain), using ICT
as the key element to make this change possible. iSalut
includes other projects like the Shared Medical Records
System in Catalonia (HC3), the Personal Health Channel
(CPS) or the model to integrate different assistance levels
(WiFIS).
The dictionary is being developed by the Office of Title
Standards
and Interoperability (OFSTI) of TicSalut foundation and many
professionals of the Catalan Health System participate as
domain experts.
The objective of the Clinical Dictionary for iSalut is to
normalize the vocabulary of the Catalan Health System
using an international standard and minimizing the
impact of its adoption. This project aims to allow semantic
interoperability between the information systems of the
different healthcare providers and to provide a homogeneous
base to represent the clinical knowledge in the Electronic
Health Record Systems (EHRS).
Material and Methods
SNOMED CT was the reference terminology selected for HC3
and is also the ontology used to build the Clinical Dictionary
for iSalut. The dictionary contains other controlled
vocabularies that are currently in use, but these
resources are mapped to SNOMED CT to guarantee the
semantic interoperability of contents.
The dictionary is conformed by different domains that are
worked as subsets of SNOMED CT, following the methodology
of creating subsets of SNOMED CT defined by the OFSTI.
The project is managed by a permanent commission that
indicates the actuation lines and the priority domains to work
on. There is a team of multidisciplinary experts for each area
of the dictionary that creates and defines all the necessary
Title
components (i.e. concepts, relationships or subsets) of
SNOMED CT in the Catalan extension of the standard.
The OFSTI distributes the international versions of SNOMED
CT and the Catalan extension with all these subsets and its
associated documentation.
Clinical Dictionary for iSalut
arius@tecnocampus.cat
12 – 18 October 2013
Beijing, China
C303
Clinical Dictionary
Part of iSalut
Based on SNOMED CT as ontology
Worked by domains and priorities
Each domain as a subset of SNOMED CT
Contains other vocabularies mapped to
SNOMED CT
Created by healthcare professionals of
the Catalan Health System
Preforms a homogeneous base for
representing clinical knowledge in EHRS.
Allowing semantic interoperability
between information systems
Results
In the first half of this year we have worked the
domains related to groups of professionals
(occupations), scales of assessment of chronic
patients and referral procedures. The second part
of 2013 we focus on rare or minority diseases and
types of clinical reports. The dictionary also includes
the revision of the spirometry test report domain
and other subsets of SNOMED CT that had been
created of allergies, immunizations and anatomic
pathology.
Conclusions
The Clinical Dictionary for iSalut is standardizing, by
priorities and domains, the controlled vocabulary used
in the information systems of the Catalan Health
System. The homogeneous base provided by this
project is allowing the unique identification of contents
to feed EHRS and Clinical Decision Support Systems
(CDSS) in research and innovation projects. The cost
of adopting SNOMED CT is significantly reduced
through the subsets-driven development and the
possibility to map the currently use vocabularies to
SNOMED CT.
The components created in the Catalan extension of
SNOMED CT are submitted to the Health Ministry of
Spain to allow its use in all the Spanish National Health
System and to consider its submission to
(International Health Terminology Standards
Development Organisation) IHTSDO.
 WHO - FAMILY OF INTERNATIONAL CLASSIFICATIONS NETWORK ANNUAL MEETING 2013

Development of a native 13606 clinical 12 -18 October 2013

Beijing, China
repository using a conceptual approach
C304
Authors: Lozano-Rubí R* ,
#† Pastor-Duran X* ,
# Canela-Soler J#

*: Medical Informatics. Hospital Clínic; #: Public Health Dep. Univ. of Barcelona;

†: Computer Science Dep. Autonomous Univ. of Barcelona

Abstract. In spite of a high investment in the development of EHRs, a new wave of resistance by healthcare professionals
is appearing. Their main complaint deals about difficulties to gather proper data to be informed about patient’s condition
and to take a decision. A new scenario will require a hard investment in knowledge representation and semantic
interoperability, as well as a deep review about the proper representation of the clinical process. We present OntoCR as an
step forward in the right direction based upon ontologies. At this moment is a full prototype, but we are near to implement
and evaluate it.

Introduction
OntoCR elements:
Computer science in healthcare has achieved 50 years of
history. During this time, one goal has been the consolidation Reference model: ISO-EN 13606-1
of a model of computer systems architecture to support the
main processes at the Healthcare facilities with several Archetype model: ISO-EN 13606-2
subsystems fully integrated in terms of information and
business process representation. But in spite to succeed in Data types: ISO 21090
the full implementation of such model, a new wave of
resistance by healthcare professionals is appearing. The Available archetypes
majority of institutions have changed only theTitletechnology
Terminologies: ICD, SNOMED CT
changing from paper, fax and phone calls to a digital
registration database. The healthcare professionals
complaint about the fragmentation of data sources and the
excessive time required to gather relevant data to be Results
informed about patient’s condition and to take a decision.
New tools have to be developed, and they will require a hard At this moment OntoCR stores data recoverable by
investment in knowledge representation and semantic any EPR in a standard format: ISO-EN 13606. Both,
interoperability, as well as a deep review about the proper the reference model and the archetype model are
representation of the clinical process, putting the s health represented into the system.
problems of the patient at the center of the IS, and allowing This system communicates with other systems using
their management and relationship with all the clinical 13606 extracts directly, without any conversion
activities. module, and accepts 13606 archetypes as
specification of clinical data structures.
Material and Methods
The presentation layer is build editing the ontology
Departure situation was OntoCRF, a product developed by which defines the graphic control for each data
our team which allows data collection for research using element, its position, and any other characteristic.
ontologies to model both, the information required by the In our solution, both, the storage and the user
researcher and the information needed to build a web user interface are obtained automatically.
interface. One of the important elements to achieve semantic
The ontologies are saved in a database specifically designed interoperability is the use of standard vocabularies. In
to store OWL ontologies. All the solution is under Liferay an OntoCR it’s possible to represent different
open source web portal, with porlets that can access directly vocabularies integrated with the ISO 21090 CD.CV
the database and extract the specific ontology, building data type.
dynamically the web pages on the fly.
OntoCR is a step forward. The metamodel of OntoCRF was
extended to represent both the reference and the archetype Conclusions
models of ISO-EN 13606 and ISO 21090 datatypes. This is
not a straightforward task, a transformation process from the
information models specified in the standards to knowledge OntoCR is a Clinical Repository. At this moment there
models to be represented with ontologies, is necessary.
Title is a β-version available
Using OntoCR it’s possible:
to use archetypes as building blocks of clinical
applications
to use standard vocabularies to identify clinical data
to communicate with other systems in a standard
way.

OntoCR is intended to support the daily clinical

practice managing the Health Problems of patients in
our Health Information System (HIS) in Barcelona city.
It would have the capability to manage the Health
Problems by physicians in real time and in an
integrated an innovative way among Family Physicians
and Specialists.
With such design based on standards it’s possible to
develop new EHRs suited for a better global healthcare
in a semantic web network

rlozano@clinic.ub.es
 WHO - FAMILY OF INTERNATIONAL CLASSIFICATIONS NETWORK ANNUAL MEETING 2013

12 – 18 October 2013
SNOMED CT and ICD-11: Beijing, China
Exploring the Value of Clinical Terms in C305
the Classification
Rita A. Scichilone, MHSA, RHIA, CCS, CCS-P
AHIMA United States
Abstract

The inclusion of SNOMED CT in the preparation of the next version of ICD adds value to the classification use in the 21st
century. All ICD entities plan to include definitions providing clear description of meaning of categories. The inclusion of
SNOMED Clinical Terms (SNOMED CT) as the foundational layer of the next version of ICD incorporates the”most
comprehensive, multilingual clinical healthcare terminology in the world” 1 into the the most widely used international
classification. As the development of ICD-11 is progressing it is important to explore, understand and appreciate the value
of the inclusion of SNOMED CT in the structure and capabilities of the classification. This poster illuminates the value of
SNOMED CT to provide additional utility to the classification and to support its role in global healthcare activities and eHealth
planning. Around the world, ”in practice the ICD has become the international standard diagnostic classification for all
general epidemiological and many health management purposes.” 2 As the recognized global standard for mortality and
morbidity statistics and its sigificant use for allocation of health expenditures, the revision with new and innovative
improvmements are welcomed. The addition of using the attributes and power of SNOMED CT is one of the significant
changes, especially for countries that have adopted SNOMED CT as the terminological standard for electronic records.
Examples of the ways the terminology brings to this century’s classification are featured here.

1. International Health Terminology Standards Development, (http://www.ihtsdo.org/snomed-ct/) accessed August 2, 2013

2. World Health Organization International Statistical Classification of Diseases and Related Problems, Volume 2, 10th
Revision, 2008

INTRODUCTION VALUE ADDED FEATURES OF SNOMED CT

The use of SNOMED CT added to the development of the

International Statistical Classification of Diseases and
Related Health Problems provides additional value to its
utility. SNOMED CT’s structure and design is more
suitable than other terminologies or classifications for
clinical documentation in the electronic health record. This
addition is an important tool around the world to support
research to enable improvement in global health.
It is important to resist the temptation to compare
SNOMED CT to ICD at the surface level in order to judge
which is “better” than the other. This would be like
comparing an automobile to a boat – the two are used for
different purposes, but each benefits from being used
“together”.
 Provides more clinical coverage by adding additional
clinical representation for additional healthcare activity
and documentation requirements
 Allows for encoding at any level of granularity as
appropriate for the clinical situation or process
 Offers multiple hierarchies important to more flexible
data entry and retrieval
 Includes attributes to support data retrieval
 Provides well defined rules to extend coverage by
combining existing concepts (post coordination)
 Facilitates direct generation of ICD codes from clinical
data enabling “capture once, use multiple times” to
save effort and reduce costs of manual data extraction
for secondary use

≠
SEMANTIC INTEROPERABILITY

Conclusions
Interoperability is a very popular word used to describe how RECENT PAPER
systems communicate with each other so that meaning is clearly
understood those dependent on and using the data Title
or
Sharing Ontology Between ICD11 and SNOMED CT:
information. Both ICD and SNOMED CT facilitate communication
for healthcare business. SNOMED CT’s structure and use of
Seamless Re-Use and Semantic Interoperability
synonyms enables clinicians to use different words to express
In August, 2013 a paper was featured in the MEDINFO
and record clinical concepts. 21st century medicine requires
2013 meeting in Copenhagen, Denmark describing e-
unambiguous and shared meaning between the sender and
health systems use of terminologies and classifications
receiver whether the system is electronic or maintained on paper.
for data representation, efficient retrieval and analysis.
For electronic systems semantic interoperability is a requirement
to enable safe and reliable communication for health care
The work continues through the WHO and IHTSDO
providers. Today’s information systems require the latest version
Joint Advisory Group (JAG). ICD-11 and SNOMED CT
of both of these standards to ensure code sets reflect current
harmonization efforts are ongoing to strengthen both
clinical knowledge.
systems for the benefit of health care around the globe.
Different words for the same meaning support communication
Rodrigues JM, Schulz S, Rector A, Spackman K,
between care providers:
Ustün B, Chute CG, Della Mea V, Millar J, Persson
KB.
Stud Health Technol Inform. 2013;192:343-6.

Access to the paper is available from

http://ebooks.iospress.nl/publication/34015
 WHO - FAMILY OF INTERNATIONAL CLASSIFICATIONS NETWORK ANNUAL MEETING 2013

Establishment of a New Scheme for 12 – 18 October 2013

Making Recommendations to the Beijing, China

Updating and Revision of ICD in Japan C306

Toshio Ogawa ,
1 Emiko Oikawa ,
2 Nobuyoshi Tani ,
2 Tomoaki Imamura1
1 Nara Medical University School of Medicine
2 The Ministry of Health, Labour and Welfare of Japan

Abstract A new scheme for making recommendations to the updating and revision of ICD has been recently established
in Japan, which is organized and managed by the WHO-FIC Collaborating Centre. All medical societies in Japan could
contribute to the ICD updating and revision under the new scheme. It would allows us to have more comprehensive and
scientific recommendations to the WHO.

Background
The International Statistical
Classification of Diseases and Related
Health Problems (ICD) has been
updated annually based on the
recommendations mainly from the
WHO-FIC Collaborating Centres to the
Updating and Revision Committee
(URC) of WHO.
There is no systematic process for
gathering recommendations from
various researchers and scientific
societies in Japan for making
recommendations to the updating and
revision of ICD.
A new scheme for gathering
recommendations from various medical
societies in Japan (hereinafter the new
scheme) has been recently established,
which is organized and managed by the
WHO-FIC Collaborating Centre in Japan.
Aim
The aim of this research is to analyse
the new scheme and to discuss the
influences of the new scheme on the
ICDTitle
updating and revision process.
Method
The new scheme was analysed based
on the interviews with the WHO-FIC
Collaborating Centre in Japan and a
number of medical societies. The
influence of the new scheme on the
ICD updating and revision process was
discussed in comparisons with the
former scheme.
<Former scheme> <New scheme>
Figure 1 Former and the new scheme for the ICD updating and revision in Japan
All recommendations from the
specialist medical societies will be
Results gathered by the WHO-FIC
Collaborating Centre in Japan and
The new scheme was established by
considered by a Scientific Committee
the WHO-FIC Collaborating Centre in
of the Centre, which consists of
Japan in collaboration with the
medical and coding experts.
Japanese Association of Medical
The recommendations will be
Sciences (JAMS), which is an umbrella
determined based on the discussions in
organization, consists of 118 specialist
the Scientific Committee. Also, The
medical societies
Title (Figure 1).
Japan Society of Health Information
JAMS refers the recommendations
Management (JHIM) provides
to the updating and revision of ICD to
suggestions to the Committee.
the specialist medical societies on
The collaboration between the
request from the WHO-FIC
WHO-FIC Collaborating Centre, JAMS
Collaborating Centre in Japan.
and JHIM will be continued through all
process until making decisions by the
WHO-URC (Figure 2).

Discussion

This new scheme would allow us to

have more comprehensive and
scientific recommendations to the
WHO Updating and Revision
Committee, compared with the old
scheme which allowed only a limited
number of researchers to make
recommendations to the ICD revision.
It would be also important to conduct
ICD revision in a systematic manners
and to clarify the division of the roles
between the WHO-FIC collaborating
Centre and medical societies.
The new scheme could contribute
to the further improvement of the ICD
in accordance with the clinical needs.
It could be a model for every countries
Figure 2 Revision Plan for 2013-14 under the new scheme in Japan Conclusions
involving the ICD revision.
 WHO - FAMILY OF INTERNATIONAL CLASSIFICATIONS NETWORK ANNUAL MEETING 2013

Identification and coding of the main 12 – 18 October 2013

Beijing, China
condition using ICD: suggested workflows
C307

Lucilla Frattura, Francesco Gongolo, Flavia Munari

Central Health Directorate, Classification Area, Friuli Venezia Giulia Region, IT WHO-FIC CC, Udine - Italy

Abstract This poster presents an Italian proposal for the systematization of current knowledge in identification and coding of a
condition introducing as well a revised workflow for the identification of the main condition.

This lively and interesting discussion on the Figure 3: Decision tree 3 - identification of
Introduction coding of the main condition in the hospital the main condition
setting, is in contrast with the lack of any
Starting from a review of the documents debate on main condition coding in the
produced at international level (1) and ambulatory setting. Our three trees (Figure NO Is the YES
taking into account the work done by the 1, Figure 2, Figure 3) are a tentative
condition
ascertained?
Italian WHO-FIC Collaborative Centre (CC) systematization that takes into account both
within an inter-regional cooperation in the hospital and ambulatory settings and is
field of children and youth neuropsychiatry compatible with some of the most common
(2), we propose a new perspective on the case-mix systems adopted in the world. NO Signs, symptoms,
abnormal results
YES

coding rules to assign the main condition. of investigations?

Figure 1: Decision tree 1 - identification of

the condition NO YES
Methods & Materials Contact with
health R codes as
Is it a single
condition?
services for main
reasons other condition
Episode of care than illness Multiple
The main condition workflow produced by conditions

the Morbidity Reference Group (MbRG) in

2010 was taken as the starting point to NO Is the condition YES Z codes as Identify the Identify the
develop three separate trees to identify the ascertained? main
condition
condition that was
the reason for
condition that
resulted in
It is the main
condition
conditions, code them, and sort out the contact with health greater resource
services use
condition, recognized at the end of the Obstetric

episode of care, primarily responsible for

the patient’s need for treatment or NO Do they YES
Of the newborn
investigation (reason for encounter vs coincide?

condition generating the greater use of NO Uncertain

diagnoses and
YES
Does the condition
resources). We adopted the perspective of symptoms? Conditions due to external
causes that resulted in greater YES
resource use depend on
clinicians who, while coding, are keen to Abnormal results
the reason for contact with
of investigations
keep their diagnosis-oriented approach. We health services?
Is there more
Identify the condition
that was the reason
Treatment of sequelae
verified the logic of our proposal by testing NO
than one reason
for contact with health
YES for contact with
health services and
it against the coding guidelines adopted in Symptoms
and signs services? that resulted in
greater resource use
Australia, Canada, Germany, and US. Contact with health services for Neither obstetric, nor of NO
reasons other than illness the newborn, nor due to
external causes, nor due

Results
to treatment of sequelae
Vaccination Social
and prophylaxis problems

Examinations Specific treatments and

Surveillance of The issue is relevant not only to achieve a
The materials we examined showed that the and procedures (attention to artif.
persons at risk
because of more standardized and comparable use of
investigations openings, appl. of prosthetic
definitions adopted in the above-mentioned (also admin. devices, surgical follow-up,
personal or family
ICD-10 but also to implement ICD-11 in the
history
Countries for hospital discharges, are purposes,
condition
rehab. treatment, organ/tissue
donors) future, since the new revision of the
different, although they all are specifications monitoring)
Contraceptive Request for International Classification of Diseases will
of the WHO definition, which has also been management,
antenatal and
advice (also on
behalf of a third easily allow customization for primary care
largely modified and updated during the postpartum
care
party)
settings. The end of the episode of care in
course of the years. the ambulatory setting is yet an open
debate.
Figure 2: Decision tree 2 - coding of the condition
Conclusions
Episode of care

We propose a systematization of the WHO

indications to code the main condition. The
NO Is the condition
ascertained?
YES
assignment rules for the main condition
should remain valid not only in hospital
Conditions due to
Obstetric Of the newborn
external causes
Treatment of sequelae
Neither obstetric, nor of settings and therefore it is highly desirable
the newborn, nor due to
external causes, nor due
the engagement of other parties in testing
Use Chapter XV Use Chapter XVI Use Chapter XX Use categories called
(O00-O99) (P00-P96) (V01-Y98) ‘Sequelae of…’
to treatment of sequelae our solution also in primary care.
NO Uncertain YES
diagnoses and
symptoms?
Symptoms and Abnormal results Acknowledgements
signs of investigations

Use ‘All the rest …’ Thanks to L. Moskal (North American WHO-FIC-

Use Chapter XVIII
(R00-R99) CC), Dr. G. Henriksson and Dr. O. Steinum (Nordic
Contact with health services WHO-FIC CC) for the help in recollection of
for reasons other than illness
background materials.
Vaccination and Social
prophylaxis problems References
Specific treatments and
procedures (attention to artif. Surveillance of 1. Rust J. Review of ICD-10 morbidity coding rules (WHO-FIC Network
Examination openings, appl. of prosthetic persons at risk
Annual Meeting, Seoul 2009 rev. Cologne 2010)
s and devices, surgical follow-up, because of
investigatio personal or VII Eye 2. Frattura L, Gongolo F, Munari F. ICD-10 implementation in the
rehab. treatment, organ/tissue III Blood & Immune V Mental X Respiratory XII Skin XIV Genitourinary
ns (also donors) family history D50-D89 Codes F Codes
VIII Ear
J Codes L Codes N Codes
health information system of the Piedmont Region (Italy) to
admin. H Codes overcome WHO multiaxial classification of mental disorders of
purposes, Contraceptive children, WHO-FIC newtork annual meeting, Beijing 2013
Request for
condition management, IV Endocrine &
advice (also VI Nervous IX Circulatory XI Digestive XIII Musculoskeletal
monitoring) antenatal and Metabolic
on behalf of a G Codes I Codes K Codes M Codes
postpartum
care
third party) E Codes
DIGITAL, MOBILE, NOW!
Use Chapter XXI

Z00-
Z13
Z20-
Z29
Z30-
Z39
Z50-
Z54
Z55-
Z65
Z70-
Z76
Z80-
Z99 Scan this to get a digital version
 WHO - FAMILY OF INTERNATIONAL CLASSIFICATIONS NETWORK ANNUAL MEETING 2013

Smartly up-to-date: an electronic 12 – 18 October 2013

tool to compare ICD, its revisions

Beijing, China

and adaptations
C308

Omar Vuattolo 1, Vincenzo Della Mea 1, Francesco Gongolo 2, Lucilla Frattura2

1Dept. of Mathematics and Computer Science - University of Udine, IT WHO-FIC CC, Udine
Central Health Directorate, Classification Area, Friuli Venezia Giulia Region, IT WHO-FIC CC, Udine

Abstract This poster introduces a tool, developed at the Italian WHO-FIC CC, to compare the possible extensions of ICD-10 with rubrics
coming both from clinical modifications and from ICD-11 beta browser, thus potentially simplifying the transition from any ICD-10 based
system to ICD-11.

Introduction Results
Since its adoption in 1948, ICD has been A first prototype (Figures 1 and 2) was developed that implements the above mentioned
progressively made suitable, through features, starting from ICD11beta (accessed through the new URI API), ICD10-CM, ICD10-
successive revisions, for grouping morbidity AM, ICD10+ (Nordic Countries).
data. Worldwide, adaptations of ICD have
also been adopted to respond to national Figure 1 – Snapshot of the proptotype window: on the left the Italian ICD-10 Vol. 1 hierarchy for adult Tcell
leukemia/limphoma; on the right, possible extensions and relative source.
requirements in terms of morbidity coding
(clinical modifications). ICD-11 is now being
developed, to be used in electronic health
records and information systems. Member
States have to use the most current ICD
revision for mortality and morbidity
statistics but one of the development goals
of ICD-11 is to contain, in its foundation
layer, all the different adaptations of ICD. A
tool to compare different adaptations and
different revisions of ICD would make
immediately available existing resources
(eg. extensions and translations) for the
design of an ICD adaptation that takes into
account classification possibilities already
explored by other national modifications and
at the same time incorporates the novelties
of ICD-11.
Methods & Materials
An informatic tool was designed to help
experts in identifying ICD-10 candidate
extensions from other available sources,
Figure 2 – Snapshot of the window for automated code assignement of selected extensions, editing of a
including ICD11 beta morbidity linearization, prompted translation from available sources, optional commenting feature
national modifications of ICD-10, and
possibly other resources such as ICD-9CM
translations and the Orphanet inventory.
Starting from ICD10 entities, candidate
extensions are prompted from the available
modifications, sorted, and when possible,
merged according to lexical rules. A web-
based interface is available for the user, that
shows an ICD10 tree browser and on its
side the set of candidate extensions,
identified as above mentioned. For each
ICD10 entity, the user views appropriate
candidates for modification, and in a second
step, can select subsets of extensions
assigning them a code. Imaging to develop
a new clinical modification or to maintain an
existing one, the set of selected extensions
can eventually be submitted via Web
Services, in form of an update proposal, to
a classification management platform and
be adopted in the respective classification.
To foster ICD11 compatibility, ICD11 entities
always appear as first choices among
extensions. Considering the possibility to Conclusions
use the tool for the Italian scenario, the
English extensions of ICD-10, were linked to The tool enables the development and maintenance of clinical modifications of ICD-10 and
the Italian translation of ICD9-CM, currently facilitates their representation as linearizations of ICD-11. Such representation simplifies
used for morbidity coding. the transition from any ICD-10 based system to ICD-11.

Acknowledgements References DIGITAL, MOBILE, NOW!

The tool has been provisionally fed with the
free online available electronic versions of
ICD-10-CM, ICD-10AM, ICD-9-CM, ICD-
10plus (Nordic Countries), ICD-11beta.
(1) Jetté N. et al. The development, evolution, and
modifications of ICD-10: challenges to the Scan this to get a digital version
international comparability of morbidity data. Med
Care. 2010 Dec;48(12):1105-10.
(2) C. Çelik, R. Jacob, T.B. Üstün Translation
Platform for ICD 11, Brasilia WHO-FIC Network
annual meeting (2012)
 WHO - FAMILY OF INTERNATIONAL CLASSIFICATIONS NETWORK ANNUAL MEETING 2013

A Common Ontology for 12 – 18 October 2013

Beijing, China
ICD 11 and SNOMED CT C309
James R Campbell, Christopher G Chute, Vincenzo Della Mea, Monica Harry,
Jane Millar, Alan Rector, Jean Marie Rodrigues, Stefan Schulz, Harold Solbrig,
Kent Spackman, Bedirhan Üstün

A joint effort to harmonize SNOMED CT with ICD 11 and other WHO Classifications has been established by a
Collaboration Agreement between WHO and IHTSDO[1] with a Joint Advisory Group, and has agreed on
principles for a Common Ontology for ICD 11 and SNOMED CT.

INTRODUCTION
ICD 11 Revision SNOMED Common Ontology Title
CT (definitions) METHODS
Architecture
SNOMED CT is a standardized • Ontological Commitment: Which kind
health terminology for of things are classified / represented
Foundation Component
health records. by
SNOMED CT formulated shares ontological core - SNOMED CT disorders
concept definitions in a with SNOMED CT and Common Ontology
a subset of SNOMED CT classes and axioms - ICD foundational component
contains additional
description logic. non-ontological - ICD linearizations
ICD11 is a multi-purpose knowledge (signs, • Analysis of hierarchical relations:
disease classification symptoms, causes, which ones correspond to subclass
extending ICD’s traditional linkage entities, relations, which ones have a different
exclusion statement)
uses for statistical reporting Title meaning
and remuneration • Analysis of exclusions and residuals in
ICD
ICD 11 has a novel two- Mortality Morbidity Primary Care … • Construction of a general framework
level architecture:
Linearizations, derived from Foundation component - Scope of common ontology (CO)
• Foundation Component
- Scope of foundation component (FC)
(FC) - a polyhierarchical
ICD 11 Revision - Characterization of non-CO nodes
collection of all relevant classes
Linkage Queries: in FC
• Purpose-specific linearizations
- Identification of linearization-
(mortality, morbidity, primary
specific nodes
care) with exhaustive and Links between
• Testbed: draft of ICD11 chapter on
mutually exclusive Foundation Component
and Linearizations cardiovascular diseases
monohierarchies

We are exploring the feasibility of a All linearization

entities are represented as
Common Ontology for SNOMED CT queries against the

and the Foundation Component[2]

Results
Title
• The best interpretations of
both SNOMED CT and ICD-
Foundation Component
classes is as Clinical
Situations [3,4], e.g.:
“Anemia” denotes the class
of situations (life periods)
of patients having anemia.
• Most of current taxonomic
links in both systems are
compatible with Situation
interpretation
• Exclusion statements are
pervasive throughout
ICD 10. In ICD 11 they will
be limited to linearizations
• Foundation Component entities
not in Common Ontology:
- Chapter headings typically using
plurals (“Diseases of…”)
- Fine-grained parts of ICD, more
specific than SNOMED CT
- Non-ontological content (signs,
symptoms, diagnostic criteria)
- Residuals (not in Foundation
Component, only in linearizations)
Common Ontology Conclusions
SELECT ?CN WHERE
Morbidity Example: (?CN SubClassOf Hypertension) • Harmonization of SNOMED CT and
Linearization
“Hypertension MINUS
(?CN SubClassOf ICD-FC feasible based on a common
excluding Pregnancy”
4
Disorders of Pregnancy)
ontology
• The SNOMED-CT / ICD
Nodes:
Common Ontology to be the
In ICD / SNOMED common ontology (CO)
IN SNOMED, not in CO core of the ICD Foundation
Residuals, only in linearizations
Component
ICD Headings
• ICD linearization codes to be
Links: linked by queries on the
is-a
Foundation Component
aggregation
(not subclasses)
• Negation in queries to be
Conclusions
“negation as failure”
• Residuals (NEC, NOS) to be
only in linearizations
Title
Outlook
• Consolidation of the ICD
revision architecture and
References integration into ICD editing
1. World Health Organization. WHO - IHTSDO Collaboration. and QA processes
http://apps.who.int/classifications/whoihtsdo/introduction.aspx
• Experimental release of the
2. Rodrigues JM, Schulz S, Rector A, Spackman K, Üstün B, Chute C,
Della Mea V, Millar J, Brand Persson K. Sharing Ontology between ICD cardiovascular chapter
ICD 11 and SNOMED CT will enable Seamless Re-use and
Semantic Interoperability. MEDINFO 2013, Copenhagen, DK. supported by common
3. Schulz, S; Rector, A; Rodrigues, JM; Chute, C; Üstün, B;
Spackman, K. Ontology-based convergence of medical ontology
terminologies: SNOMED CT and ICD 11. eHealth 2012, May 10-
11, 2012; Vienna, AT. • Extension to other WHO
4. Schulz, S; Rector, A; Rodrigues, JM; Spackman, K Competing
classifications
Interpretations of Disorder Codes in SNOMED CT and ICD. AMIA
Annu Symp Proc. 2012; 2012:819-827
 WHO - FAMILY OF INTERNATIONAL CLASSIFICATIONS NETWORK ANNUAL MEETING 2013
Leveraging ICD Data to
Improve Global Health
American Health Information Management Association
Abstract New and improved classification systems are better able to meet changing health information demands,
allowing improved capture of information about the increasingly complex delivery of healthcare. This poster explores
some of the exciting new ways in which the advent of new classifications leads to expanded use of coded data to
improve the care of individuals and populations and the effectiveness and efficiency of healthcare delivery.
Introduction
An increasingly global, complex, and
electronic healthcare environment
and an accelerating body of
knowledge regarding diseases and
medical advances have led to ever-
expanding demands for richer health
information. Accurate and detailed
information about clinical conditions
and the services rendered are
essential to ongoing management of
care as well as refinement of
healthcare delivery systems.
Adoption of electronic health record
(EHR) systems and interoperable
health information networks require
sophisticated classifications for
summarizing and reporting data. The
demand for better, more detailed
healthcare data is also driven by
emerging or expanding data-
dependent initiatives such as value-
based purchasing, accountable care,
quality measurement and patient
safety programs, and changing
healthcare reimbursement models.
Close alignment between
classifications and terminologies will
allow rich healthcare data to be
extracted at the level of aggregation
or granular detail needed for the
intended purpose.
Public Health
Modern code sets will permit a level
of precision that hasn’t been possible
in the past, enabling more meaningful
data and analytics and greater health
intelligence.
Public Health
As a result of both more detailed ICD
data and the widespread adoption of
EHRs, ready access to collective data
worldwide would allow for early
exchange of public health information
to identify disease outbreaks and
bioterrorism events, allowing action
to be taken more swiftly.
Research
Better understanding of diseases and
injuries will lead to improved
prevention or mitigation strategies.
Clinically robust algorithms to treat
chronic diseases and track outcomes
of care can be designed. Greater
detail offers the ability to discover
previously hidden relationships or
uncover phenomena such as incipient
epidemics early.
Sue Bowman
Quality of Care
Data can be used in more meaningful
ways to better understand
complications, monitor patient
safety, track care outcomes, and
design clinically robust algorithms.
Best practices across the healthcare
industry can be identified, resulting
in the sharing of more effective and
cost-conscious treatments targeted
to the patient’s disease state or
Title
Research
clinical condition. Individuals’ health
status can be monitored over time as
well as the effectiveness of health
interventions.
Research
Better Implementation
ICD data will permit improved
Guidance
Implementation
identification of patients for disease
Guidance
management programs and more
effective tailoring of these programs
to meet individual patient needs,
thus improving patient outcomes,
patient satisfaction, and lowering
healthcare costs.
Effectively practicing population
health management will require the
ability to innovatively assess and
stratify patient demographic and ICD
data into categories, such as patients
who are well, at risk, or have chronic
conditions.
Consumer Engagement
Coded data can be used to provide
consumers with the ability to
compare quality and cost across
healthcare providers. With better
data will come an expanded ability to
educate consumers on costs and
outcomes of treatment options.
Increased patient understanding and
Title
involvement in their healthcare will
improve population health and
decrease healthcare costs.
Fraud and Abuse
A classification with less ambiguity,
more detail, and improved logical
structure and organization will help
reduce opportunities for fraud and
improve fraud detection capabilities.
Increased specificity will make it
easier to compare codes with clinical
documentation, check for consistency
between data elements, and check for
illogical combinations of diagnoses.
Improved logic and increased
specificity will facilitate the
development of sophisticated tools for
fraud detection.
12 – 18 October 2013
Beijing, China
C310
Reimbursement
Administrative Efficiencies
Modern classifications provide more
detail for determining payment policies
and reimbursement rates under existing
reimbursement programs as well as for
new reimbursement systems based on
quality measurements and outcomes,
value-based purchasing, and such as
accountable care organizations and
Administrative
bundled payments. Efficiencies
Administrative Efficiencies
Increased administrative efficiencies and
lowered costs will result from increased
use of automated tools to facilitate the
code process, fewer rejected and
improper reimbursement claims, fewer
coding errors, increased productivity, and
less reliance on manual review of medical
records for audits, research, and other
purposes.
Organizational Performance
Better data will support providers’ and
payers’ efforts to improve performance,
create efficiencies, and contain costs.
Considerable cost savings can be
realized through more accurate trend
and cost analysis. Payers will better be
able to forecast healthcare needs. Both
payers and providers will be able to
more effectively monitor service and
resource utilization and patient
outcomes.
Conclusion
While the introduction of new
classifications is costly and disruptive to
Conclusions
mortality and morbidity statistics, it is
essential to meet the expanding
demands for health information.
A classification system designed for the
21st century would maximize
investments in EHR systems because it
would facilitate data retrieval at the
desired level of detail and lead to the
expanded use of technologies to improve
data quality and efficiency of data
collection, such as computer-assisted
coding and natural language processing.
The information a healthcare
organization collects needs to be
managed as a business asset to ensure
it is trustworthy and actionable, because
in the future, accurate and timely
information will be the greatest asset to
drive successful healthcare
organizations.
 WHO FIC NETWORK ANNUAL MEETING

ICD Morbidity Coding Rules for Main Condition,12 – 18 October 2013

Beijing, People’s
Diagnosis Timing, Numbers of Diagnosis Fields Republic of China
H Quan,1 S Droesler,2 V Sundararajan,3 C311

L Moskal, W Ghali, H Pincus

Quality & Safety Topic Advisory Group (Q&S TAG)
1University of Calgary, Canada 2Niederrhein University of Applied Sciences, Germany 3University of Monash, Australia
Abstract Hospital-based medical records are abstracted to create International Classification of Disease (ICD) coded discharge
health data in many countries. We describe how the “main condition” in ICD data is currently defined across countries, the impact
of these definitions on research and analysis, and propose recommendations for international harmonization.
compared to six countries that did not.
Introduction Results WHO Collaborating Centre
Our analysis showed that 89-90% of
The international standardization of Main Condition. The “main condition” central venous catheter infections and
ICD has been compromised by is not defined in a consistent manner 97-99% of retained foreign bodies and
inconsistency among countries in ICD internationally. Of 30 countries accidental punctures/lacerations were
coding rules. One notable issue is the surveyed, 17 employ a “reason for captured within 15 secondary diagnosis
coding rule for main condition (i.e., the admission” coding rule as the basis for fields.
principal or primary diagnosis). Some the main condition, while 13 countries
employ a “resource use” coding rule. Figure 1. Mean Proportions of Secondary
countries use a ‘reason for admission’ Diagnoses Counts.
definition while other countries use a
‘resource use’ definition. The latter Table 1. Main condition definition in ICD
leads to assignment of a different coded health data.
diagnosis as the main condition when a Resource Reason for Coding
patient has been admitted for one Country
Use Admission System
condition but then suffers a significant Australia X ICD-10
complication that requires a Belgium X ICD-9
significantly prolonged hospital stay Brazil X ICD-10
and use of considerable resources. X X ICD-10
Canada [Province of
Other key areas of international Quebec]
inconsistency include: China X ICD-10
Denmark X ICD-10
1)the number of available diagnosis Finland X ICD-10
fields in hospital records (ranging from France X (2009) ICD-10
as few as two to as many as an infinite Germany X (2001) ICD-10
number of fields); Iceland X ICD-10
Ireland X ICD-10
2)the availability of potentially powerful Italy X ICD-9-CM
diagnosis timing indicators in Canada, X (2001) ICD-10
Japan
the US, and Australia, with potential for
Latvia X ICD-10
more widespread implementation in ICD-9-CM/
ICD-11; and Netherlands X
ICD-10
New Zealand X ICD-10
3)exploration of approaches to Nicaragua X ICD-10
diagnosis clustering that would permit Norway X ICD-10
better coordination and linkage of Diagnosis timing indicator.
Portugal X ICD-9
diagnostic concepts in hospital A diagnosis timing indicator, also known
Singapore X ICD-9
discharge records. as “present on admission (POA),”
South Africa X ICD-10 improves the ability of ICD coded hospital
South Korea X (2012) ICD-10 discharge data to support outcomes
We described these coding rules in ICD Spain X ICD-9
data across countries and the impact of research and the development of quality
Sweden X ICD-10
these rules on research and analysis, and safety indicators. The importance of
Switzerland X ICD-10 POA reporting is that enhancing risk-
and propose recommendations for
international harmonization of these Thailand X ICD-10 adjustment with POA-corrected
coding rules. United X ICD-10 covariates may have a substantial impact
Kingdom on estimates of hospital performance,
United States X ICD-9-CM especially for conditions and procedures
Methods Venezuela X ICD-10 that often involve acutely ill patients. An
Note: Brackets indicate date changed. advantage of POA reporting is its ability
A literature review was completed to exclude relevant events with a higher
and two surveys were conducted Secondary Diagnosis Field.
Patient safety indicators (PSI) rely on likelihood of being present on admission
among a subset of member countries and focus on complications only that
of the World Health Organization relevant events coded in the secondary
diagnosis field. There is a positive occurred during hospitalization.
Family of International Classifications
(WHO-FIC) Network and the association between PSI rates and the
Conclusion
Organisation for Economic Co- mean number of secondary diagnoses.
operation and Development (OECD). We investigated the possible
association by testing whether three We propose a method of harmonizing the
In the survey, we asked key international definition to enable
stakeholder respondents to indicate: countries that expanded their data
systems to include more secondary researchers and international organizations
"What is the current definition for the using ICD-coded health data to aggregate
main/most responsible diagnosis in diagnosis fields showed increased PSI
rates or compare hospital care and outcomes
ICD data [in your country]?" across countries in a consistent manner.
 WHO - FAMILY OF INTERNATIONAL CLASSIFICATIONS NETWORK ANNUAL MEETING 2013
Field Testing ICD-11
in Canada
Hude
Nenad
for the Quality & Safety TAG
1University of Calgary, Canada 2World Health Organization, Geneva
RATIONALES
The wide use of ICD data has been facilitated by
their: Verdana 36 Bold
 readiness to be analyzed,
 wide geographic coverage and
 relatively complete ability to capture episodes
of patient contact with the health system.
The use of ICD data for any purpose is based on the
assumption that ICD data provides valid information
on diagnoses and clinical services.
Title
However, errors could occur in the process of
creating ICD data, due to incomplete information in
ICD-11 and coders’ misinterpretation of ICD
diagnoses.
ICD-11 Advancements
 ICD-11 has more codes than previous versions and
elaborates diagnoses and symptoms more in-depth.
 ICD-11 codes are linearized— providing the ability to
extract ICD-11 codes for a specific task.
 Concise and structured definitions and guidelines are
available for ICD-11 codes.
 ICD-11is a single standard coding system that will
replace various country specific ICD-10 versions.
Systematically testing the ICD-11 before its use will
identify improvement paths, reduce errors and
increase its consistency. The ICD-11 should be
easy to use (feasibility), generate same results in
the hands of all users (reliability) and add value
(utility). It will fit multiple purposes and be
comparable with ICD-10.
OBJECTIVES
Reliability of ICD-11 coding among coders.
 Comparability between ICD-10 and ICD-11 in
defining morbidities.
 Data quality improvement generated from ICD-
11 compared with data coded with ICD-10.
 Impacts of coding rules on condition identification
and grouping.
 Utility of ICD-11: experience of ICD-11 use for
coding morbidity.
12 – 18 October 2013
Beijing, China
C312
Quan,
1 Kostanjsek 2
METHODS
ICD-10 Data
Title
We will identify inpatients discharged between January 1,
2013 and June 30, 2013 in the teaching hospitals in
Calgary, Alberta, Canada. The identified discharge
records will be stratified by hospital site. Medical
discharge records will be purposively selected from
hospitals. The selected records will have conditions,
symptoms or injuries that reflect changes in ICD coding
guidelines. These data have already been coded using
ICD-10-CA.
ICD-11 Data
To create the new ICD-11 database, coders who have
ICD-10-CM coding experience at these hospitals will re-
code the charts following the ICD-11 coding guidelines.
The coders will be blinded to the original ICD-10-CA
codes assigned in each hospital.
Chart Review Data
Corresponding patient charts for the selected discharge
records will be located using a combination of the patient
chart number and admission identification number that
are unique to admissions at each hospital.
Two clinically trained chart reviewers will undergo
training in the data extraction process. Once the two
reviewers reach substantial agreement, they will extract
data independently. The reviewers will examine the
entire chart to define conditions.
Charts
Data Data Data
ICD-10 ICD-11 Chart
Conclusions
Title
Coder A Coder B Coder C Coder D
ANALYSIS
We will describe samples and analyze the data using a
Kappa score for agreement among coders, and
sensitivity, specificity, negative predictive value and
positive predictive value for data validity. McNemar’s
test will be used to compare the sensitivity of ICD-11
data and ICD-10 data relative to chart review data for
detecting conditions. The specificity of the ICD-11 and
ICD-10 data relative to chart data will then be compared
to cases without the condition.
Experience of the utility of ICD-11 will be described in
text. Strategies of improving ICD-11 coding and coding
guidelines will be proposed.
 WHO - FAMILY OF INTERNATIONAL CLASSIFICATIONS NETWORK ANNUAL MEETING 2013

12 – 18 October 2013
Quality & Safety TAG Beijing, China

C313

Hude Quan, Danielle Southern

for the Quality & Safety TAG

OBJECTIVES ACTIVITIES
Mapping of existing patient safety
The Quality and Patient Safety TAG is indicator
We have held meetings in both New
chargedVerdana
with reviewing ICD-10, ICD- York, NY and Washington, D.C., US Title
36 Bold We have begun a mapping exercise,
10CM and progressive drafts of ICD-11
to inform the development of the ICD- whereby we have attempted to map
 Reviewed the status of discussions
11, focusing on identifying practical the Calgary PSI list, the International
around coding rules (main condition,
modifications for ICD 11 drafts that ICD-10 AHRQ PSI list as well as Patient
diagnosis timing, coding field).
would enable better measurement of Safety concepts in ICD-11-Beta.
 Reviewed chapter 19&20 content
quality and safety. and associated clustering mechanisms
Code-recode testing
and presented these concepts in
Ultimately, an enhanced classification emails to WHO.
system will permit expanded use of OBJECTIVES
 Prepared to undertake a granular
coded health data for large-scale review of the content in chapters 1-20
quality and safety surveillance in  To assess, from a healthcare
and willTitle
devise a committee work plan
health care systems internationally. leader’s perspective, the utility of
to do this.
patient safety information encoded
TASKS using the following classification
Progressing Manuscripts
systems: AHRQ Common Format,
 Horizontally crossing all ICD-11 WHO-ICD 10-CA, and WHO-ICD11
The editorial team of the International
chapters to advise on optimizing the (Beta)
Journal for Quality in Health Care is
entire classification’s content,  To evaluate the inter-rater reliability
welcoming a series submission (i.e.
structure and coding rules for of raters classifying patient safety
intermittent submission of papers as
enhanced application in both existing events
they are completed).
versions.  To determine the face validity of
 Developing an inventory of existing •Main Condition event classification
quality of care and patient safety •Number of diagnoses fields  To assess the coding practice for
indicators and potentially novel quality •Timing of diagnosis classifying patient safety events
and safety indicators. • Overview TAG
 Assessing potential uses of ICD-11 • New PSI project The proposed study will achieve its
for health services, quality and patient- • 19&20/concepts objectives by reviewing detailed
centered outcomes research. descriptions of patient safety events
 Reviewing and critiquing the ICD-11 Progressing Field Trials detected from the prospective
alpha draft from the perspective of the surveillance of hospitalized internal
quality and safety use case. The QS-TAG has devised a matrix medicine patients in 5 hospitals in
 Providing input to the ICD-10CM model for considering potential ICD-11 Ontario and Quebec, Canada. We
refinement, implementation and field trials. The matrix categorizes monitored 1346 patients from
maintenance in the US, and ultimately cross-tabulates topic areas (e.g., admission to the service until
designing field trials for the beta validity of coded concepts, disposition. Trained observers (who
version of ICD-11. completeness of capture of critical were either MDs or RNs) performed the
patient safety and quality concepts, following tasks daily: staff interviews,
reliability and feasibility of various medical record reviews, and clinical
MEETING ATTENDEES coding rules, opinions of stakeholders rounds observations, to detect
on various issues) against the occurrences suggestive
Conclusions of adverse
Australia: James Harrison, Vijaya methodologies that would be used for events or potential adverse events.
Sundararajan the field trials (i.e., code-recode Once an occurrence was identified, the
Title clinical reviewer described it fully so
studies using real medical records,
US: Marilyn Allen, Chris Chute, Ginger coding studies assessing completeness that it could undergo multidisciplinary
Cox, Donna Pickett, Harold Pincus,, of capture of key safety/quality review. During weekly review sessions,
Patrick Romano, concepts, surveys of stakeholders, events were categorized into adverse
Brigitta Spaeth-Rublee, heuristic evaluations of ICD-11 on events, potential adverse events, and
various user interfaces, etc). non-events. We, thus, identified 546
Canada: Susan Brien, Alan Forster, occurrences that represented adverse
William Ghali, Yana Gurevich Survey events or potential adverse events.
Each of these occurrences is described
Switzerland: Bernard Burnand, With the overriding goal for the TAG (& in detail, including the patient’s
Lori Moskal, Hude Quan, thus the WHO) to collect info on user background medical conditions, the
Danielle Southern needs from ICD-11 in advance of the factors leading to the occurrence,
next TAG meeting (in September) to response to the occurrence by the
France: Cyrille Colin inform ICD-11 refinements. We have healthcare team, and how the
developed a survey for the field trial. patient’s condition was affected.
Germany: Saskia Droesler http://fluidsurveys.com/surveys/qs-
who/quality-and-safety-in-icd-11/ ACKNOWLEDGEMENT
WHO: Nenad Kostanjsek, Bedirhan Q&S TAG was funded by AHRQ and
Ustun. Canadian Institute of Health Canadian
Patient Safety Institute (CPSI).
 WHO - FAMILY OF INTERNATIONAL CLASSIFICATIONS NETWORK ANNUAL MEETING 2013

12 – 18 October 2013
Leveraging ICD11 Field Beijing, China
Trials for ICD10 Evaluation C314

Marjorie S. Greenberg, Mea Renahan, Lori Moskal,

Sam Notzon and Donna Pickett

Abstract The upcoming ICD-11 field trials present an ideal opportunity to ask participants several standardized questions,
as part of the Basic Questions, about whether their countries have fully implemented ICD-10 and its updates for both
mortality and morbidity classification, and, if not, what the barriers are to implementation. Lessons learned from this
information and other WHO-FIC Network studies and activities can be applied to ICD-11 implementation and updating.
Background
In the oft-quoted and misquoted
words of the Spanish American
philosopher, George Santayana, “Those
who cannot remember (or learn from)
the past (or history) are doomed (or
condemned) to repeat it”. This
admonition comes to mind when
considering past attempts to evaluate
the implementation of ICD-10 and its
updating process. In 1997, the Heads
of the Collaborating Centres agreed
that no consideration should be given
to ICD-11 until after an evaluation of
the updating mechanisms for ICD-10
had been carried out and the results
considered by WHO and the Centres.
The Head of the North American
Collaborating Center (NACC) submitted
a paper to the 1999 Collaborating
Centres meeting, “Approaches for
Evaluating the Updating Process for
ICD-10”; this paper was accepted in
principle and referred to the Update
Reference Committee.
In 2004, the Heads of the United
Kingdom Centre and NACC presented a
paper at the annual WHO-FIC Network
meeting on “A Strategy for Evaluating
ICD-10 Implementation and the
Updating Process”. The paper put
forward proposals for systematically
identifying the lessons to be learned
through formal evaluations of how well
the key objectives of the classification
were supported by the main activities
involved in implementation and
updating and that these findings should
guide plans for revision of ICD-10 and
implementation of ICD-11.
In February 2005, a preliminary
consultation on the evaluation was held
at NCHS, with participation from the
UK Centre, Australian Collaborating
Centre, Canadian Institute for Health
Information, PAHO and WHO. The
Heads of the UK and NACC
subsequently submitted a status report
to the 2005 Annual meeting, with a
study design and two proposed
questionnaires. Although some
components of the evaluation have
been conducted through Education and
Implementation Committee surveys,
the Update and Revision Committee
and the Regional Networks, no
systematic evaluation of the facilitators
and barriers to implementing ICD-10
and its updates has ever been
undertaken.
What we know
Implementation of ICD-10
•Information in the current ICD-10
implementation database is
incomplete and outdated. The Dutch
Centre, working with WHO and the
Education and Implementation
Committee (EIC), is piloting a new
web-based application that will be
discussed in Beijing at the 2013
meetings
Title
•Approximately 100 countries
responded to a survey fielded by the
Education Committee in 2004; the
results reported on implementation of
ICD-10, barriers to implementation,
mortality and morbidity coding and
coder education and support.
•The Pan American Health
Organization and the Asia Pacific
Network have gathered
implementation information since
2005 and reported at meetings.
•Several countries have presented
papers on implementation
experiences and challenges.
•EIC has developed an
Implementation Checklist
Implementation of Updates
•Questions about awareness and
implementation of updates are
included in the current ICD-10
Implementation Database, but the
information is incomplete and
outdated. Additional questions will be
included in the new Database.
•The Update and Revision Committee
(URC) prepared papers in 2004 and
2006 that addressed accomplishments
and challenges of the ICD-10
updating process
Title and identified
clinical updating projects that had
been beyond the capacity of the
Committee.
•The Regional Networks have
documented uneven adoption of ICD-
10 updates and have identified
challenges, such as lack of awareness
and lack of resources (financial,
translation, nosology).
•Even countries that participate
actively in the WHO-FIC Network
updating process have not always
been able to implement the updates
in a timely manner (e.g., the U.S. in
its automated mortality coding
system).
•There is considerable variation in
how frequently countries update their
national versions of ICD-10.
Title
What we’d like to know
The upcoming ICD-11 field trials
present an ideal opportunity to ask
participants several standardized
questions, as part of the Basic
Questions, about implementation of
ICD-10, awareness of the ICD-10
updating process and whether their
countries have implemented the
updates in their national versions of
ICD-10 used for mortality and
morbidity classification. One of the
two main purposes of the proposed
field trials is to ensure comparability
between ICD-10 and ICD-11. Yet all
of the basic questions currently
proposed are about ICD-11. It is
critical that several questions are
included on current use of ICD-10 and
its updates because ICD-10 is
considered a bridge to ICD-11. This
can supplement similar questions that
have been included in the new WHO-
FIC Implementation Database and
raise awareness about the updating
process. Field trials often include
persons and countries not typically
involved in the WHO-FIC Network and
offer a “captive audience” for
collecting parsimonious information.
They may also provide an opportunity
for focus groups. By identifying
barriers to implementation, it may be
possible to develop mitigation
strategies that will improve the
adoption of ICD-11 and the uptake of
ICD-11 updates and promote more
consistent data on a global basis.
Questions could address demand
Conclusions
for updates, limitations of the
updating process, demand for national
modifications, achievements of the
updating process, barriers to full
implementation of the updating
process and impact of updates and
timing on clinical, statistical and other
data.
It also would be informative to
evaluate whether the clinical areas
that could not be addressed in the
ICD-10 updating process have been
successfully addressed in ICD-11,
because this was one of the major
rationales for embarking on ICD-11.
These areas have been documented
by URC and by the Morbidity
Reference Group in several papers.
Copies of papers cited are available
from the first author. Systematic
synthesis of information in these and
related papers is recommended.
 WHO - FAMILY OF INTERNATIONAL CLASSIFICATIONS NETWORK ANNUAL MEETING 2013

Tool Support for Collaborative 12 – 18 October 2013

Beijing, China
Terminology Authoring: C315

The Stanford WHO Collaborating Centre

Mark Musen MD PhD, Tania Tudorache PhD, Csongor Nyulas MS, Samson Tu MS
Abstract The Stanford WHO Collaborating Centre provides modeling support and a common software infrastructure for
developing and maintaining WHO classifications. Our foundational software includes iCAT, a Web-based tool that
Title
enables subject-matter experts to browse, review, edit, and comment on draft versions of WHO classifications. The
iCAT platform additionally allows terminology authors to access standard terminologies hosted in the BioPortal
repository of the United States National Center for Biomedical Ontology. Through BioPortal, developers who use iCAT
can link their evolving content to terms in reference sources such as SNOMED CT. Additional work enables WHO to
review the overall process of terminology development through visual presentations that show areas of evolving
classifications that are undergoing rapid change, or that may need additional attention. Future work involves support
for modeling and implementing post-coordination in ICD-11 and ICTM, as well as the design and development of
methods for enhanced community engagement for commenting on the ICD-11 Beta draft.

iCAT for ICD Revision iCAT for ICTM

iCAT provides support for: iCAT features (continued):

 Real-time collaboration Simple form-based content acquisition
 Notes and discussions Full change history log
 Export to MS Excel User role management
 Management of hierarchy and linearizations Multi-lingual labels, and much more …

System Components iCAT Analytics

WHO’s Nightly iCAT Edits

ICD-11 Public Browser Dumps http://icat.stanford.edu/
http://apps.who.int/classifications/icd11/

Terminology author

Synchronize Publish
Notes ICD-11 revisions
Browses and
Adds Comments ICD-11
Content Model
NCBO BioPortal
http://bioportal.bioontology.org/

Browses and  Concept nodes size ~ Number of changes  Timeline of the number of changes
Adds Comments  Edges = Parent-child relations  Chart depicting contributions of authors
 Blue nodes = Ready for public comment  Lists of parents and children
Terminology user
 Yellow nodes = Work-in-progress
 Red nodes = Need much more work.

National Center for Biomedical Ontology BioPortal

Repository of >350 ontologies and terminologies

 Web services for accessing ontology terms
 Public comments
 Mapping to multiple ontologies
 Index concept usage in resources
 WHO - FAMILY OF INTERNATIONAL CLASSIFICATIONS NETWORK ANNUAL MEETING 2013

Crowdsourcing ICD-11 12 – 18 October 2013

Beijing, China
Sanctioning Rules C316

Samson Tu MS, Vincent Lou, Csongor Nyulas MS, Tania Tudorache PhD,
Robert J. G. Chalmers MB FRCP, Mark A. Musen MD PhD
Abstract ICD-11 is a post-coordinated classification system, where end users can combine disease codes with qualifiers to
form detailed descriptions of diseases. To prevent nonsensical combinations of terms, we need to find sanctioning rules that
only allow those sensible combinations to be formed by end users. In this pilot study, we use crowdsourcing methods to
find sanctioning rules between diseases and anatomic locations.

ICD-11 Post-Coordination Sanctioning Rules

Title
Sanctioning rule:
1
2
(disease, anatomic location)
3
1
•The disease can occur at the specified
anatomic location or its descendants.
3
2

3 3 2 1 1
1
2
3
The following turker-selected anatomic
locations (dark green and light green):
Pre-coordination Post-coordination Elbow
Experiments
Head and
•Enumerate all •Combine codes to Neck

codes that may describe a complex Upper

Forearm Finger
Experiment settings:
represent complex concept Extremity Hand Palm •11 disease branches; 209 diseases total
concepts •Smaller
Trunk
Foot Forefoot
•Number of anatomic locations: 335
•Large classification classification Lower •Average 19 diseases per branch
Extremity Lower Leg Midfoot
•Have to specify Sample branch output:
Hip Heel
sanctioning rules to Disease Gold Turkers’ Sensitivity Specificity
permit sensible generate the sanctioning rules:
standard responses

combinations
(Palmoplantar keratodermas, Hand) Hordeolum Eyelid Eyelashes 0.0 0.99
(Palmoplantar keratodermas, Foot)
Hordeolum Eyelid,
internum
Eyelid
Conjunctiva 1.0 0.98
Method Hordeolum Eyelid margin,
internum Eyelid Canthus, 0.64 1.0
lower eyelid Lower eyelid

Elbow
Head and Sample summary statistics per branch:
Palmoplantar Neck
Forearm
keratodermas
Upper Root concept of Number of Average Average
Extremity Hand ICD-11 branch diseases Sensitivity Specificity
Trunk
Hereditary Acquired
Foot An anatomical location is
Hordeolum 3 0.55 0.99
palmoplantar palmoplantar
keratodermas keratodermas Lower selected if >= 5/10
Extremity Lower Leg Turkers picked it. Experimental Results:
Hip Root concept of Number of Average Average
ICD-11 branch diseases Sensitivity Specificity
Dermatoses of the
scalp 34 0.97 0.97
Elbow
Infective disorders of
Head and
Neck the external ear 20 0.66 1.0
Forearm Genetic syndromes

…
Palmoplantar
Title keratodermas Upper
Extremity Hand
with abnormalities of
the hair shaft
11 1.0 1.0
Certain specified
Trunk
disorders of external 12 0.86 1.0
Hereditary Acquired ear
palmoplantar palmoplantar Lower
Infectious disorders of
keratodermas keratodermas Extremity
eyelid 27 0.52 0.92
… repeat above steps until the sanctioning rules are found, Disorders of lips 33 1.0 0.99
then repeat for child diseases Disturbances of oral
epithelium 10 0.4 0.97
Head and
Neck Lichen planus and
12 0.92 0.92
Palmoplantar
keratodermas lichenoid reactions of
Upper oral mucosa
Extremity Hand Non-infective erosive

Trunk
and ulcerative disorders 30 1.0 0.95
Hereditary Acquired of oral mucosa
Foot Acquired disorders of
9 0.58 0.98
palmoplantar palmoplantar
keratodermas keratodermas Lower eyelashes
Extremity Inflammatory disorders
of eyelid 11 1.0 1.0
Number of Average Average
branches Sensitivity Specificity
Evaluation Metric
Title
11 0.81 0.97

Gold Standard: Elbow

Cost Analysis:
Head and
(Palmoplantar keratodermas, Hand) Neck
Finger
•We asked 641 questions; we received 6410
Forearm
Upper responses; we paid 3296 bonuses
(Palmoplantar keratodermas, Foot) Extremity Hand Palm
•We paid Turkers 4 cents per response and 2
Trunk
Foot Forefoot cents per bonus. Amazon Mechanical Turk
Suppose Turkers indicate: Lower charged 0.5 cent for each transaction.
Extremity Lower Leg Midfoot
(Palmoplantar keratodermas, Hand) •Cost to acquire sanctioning rules per
(Palmoplantar keratodermas, Lower Leg)
Hip Heel
disease = (6410*0.045+3296*0.025)/209 =
$1.77 USD
Then: •Average time per response: 55 seconds
True Positive False Positive True Negative False Negative Sensitivity Specificity •Total time for getting all responses: 82h
3 1 7 4 0.43 0.88 •Total number of Turkers: 135
 WHO - FAMILY OF INTERNATIONAL CLASSIFICATIONS NETWORK ANNUAL MEETING 2013
Universal Health Coverage: Information and Innovation 12 – 18 October 2013
Beijing, China
The 7 steps to co-create global knowledge
in healthcare on Twitter. C317

The ICD example

Authors: de San Pedro M, Canela-Soler J, Pla M, Martínez J, Dedeu A, Pujante JA, Serrano J,
Alzaga X, Tognetta F, Fontanet G, Fortes M, Loyola-Elizondo E, Argimon JM, Garcia-Cuyàs F.
Location: Barcelona
Abstract
Abstract The microblogging service Twitter and its conversational interaction features combined with third parties tracking platforms, are
contributing to consolidate and deliver universal knowledge, as is shown in “The International Classification of Diseases (ICD)
example”. The process seems to be aligned with the co-creative paradigm and could be the framework to deal with new healthcare challenges. The
assumption is, that there aren’t right answers, but evolving conversations open to discussion. The seven step process towards global knowledge in
healthcare proposes the stages that should be considered in a global strategy in healthcare on Twitter: Community, Connection, Co-design,
Content, Conversation, Consolidation and Co-creation. Unleashing the power of Twitter and co-creative processes are key to meet ICD challenges.

The 7 step process towards global knowledge

Introduction
Introduction
Can Twitter be helpful to co-create new
solutions for healthcare challenges?
If so, we should also, as a result, provide
appropriate answers to new research
questions: ·What community of individuals
do we need to connect to solve a problem?
·What new interactions will community
members want to engage in on Twitter to
design a solution? ·What valuable
professional experiences will the members
get out of these interactions? ·What value
will this new set of experiences generate for
the community members, creating a win for
all parties?
TWITTER PARTICIPANTS
Methods & Materials
in healthcare on Twitter is based on the
Methods
framework of collaborative networks under
 Conversation analysis the co-creative paradigm. The assumption
 Problem solving in public health is, that there isn’t a right answer, but an
 Open innovation evolving conversation open to discussion:
1. Community: Engaging in open conversations.
 Socratic method
2. Connection: Taking advantage of Twitter.
 Mixed Methods in the social and
3. Co-design: Supporting a culture of innovation.
behavioral research
4. Content: Contributing with relevant information.
Materials 5. Conversation: Adopting healthcare hashtags.
 Twitter in Healthcare: Conversations & 6. Consolidation: Harnessing healthcare content
Hashtags: HIMSS (http://bit.ly/Y66xZV). curation platforms such as “Symplur”.
 The #ICD10 hashtag: Symplur 7. Co-creation: Experiencing meaningful value.
(http://bit.ly/18HvI8T).
PROCESS OF CO-CREATION
 Community-Powered Problem Solving:
HBR.org (http://bit.ly/14BdMOh).
 Communities of Interest: University of
Colorado (http://bit.ly/ZdwMiM)

 Co-Creation of Content to promote

Learning, Activism and Advocacy: Nursing-
Informatics (http://bit.ly/1c0gFfi).

 Universal Health Coverage: WHO

(http://bit.ly/13dav2s).

Results

THE ICD EXAMPLE

Conclusions & Recommendations

Conclusions
· A precondition for successful adoption of
Objectives
Twitter in healthcare is a positive attitude
• Identify key attitudes and values for and openness toward its innovative
innovative Twitter participants in information potential.
healthcare. • The systematic use of ICD hashtags and
• Highlight innovative #ICD conversations healthcare content curation platforms, such
as Symplur, contribute to the extension of
in terms of open information and its
universal health knowledge.
impact on universal health coverage.
Recommendations
• Design a coherent process to determine
the cycle of co-creation in healthcare on • Unleash the power of Twitter and co-
creative processes to meet ICD challenges.
Twitter.

Acknowledgements: Virtual Nurse, Official College of Nursing in Barcelona (www.infermeravirtual.com); Virtual Doctor, DKV Medical
Insurance (www.dkvseguros.es); The Catalan Agency for Health Information, Assessment and Quality (www.aatrm.net) and
UniversalDoctor (www.universaldoctor.com) collaborate on this project. This work is conducted within the framework of the doctoral
program in Health, Wellbeing and Quality of Life at the University of Vic – Barcelona (www.uvic.cat).
 WHO - FAMILY OF INTERNATIONAL CLASSIFICATIONS NETWORK ANNUAL MEETING 2013

ICD URI Model

12 – 18 October 2013
Beijing, China

and Supporting Web Services C318

Celik C.1, Della Mea V. 2, Noselli M. 1,3, Vuattolo O. 2

1CTS, World Health Organization, Geneva, Switzerland; 2 Italian Collaborating Center
and University of Udine, Italy; 3 University of Freiburg, Switzerland
Abstract The present poster briefly describes the ICD URI model that provides unique identifiers to WHO classifications
and their entities, and the supporting Web Services that provide classifications and entities in computable format for their
adoption into information systems. The poster anticipates a forthcoming guideline regarding ICD-11 and ICD-10, that could
be adopted also for the other classifications.

Introduction Scope Service URIs: linearizations

In the past, WHO-FIC discussed
methods for timely availability of
classifications to information systems,
as well as more unique identification
schemes for classifications and
classification entities.
Two main concepts are at the basis of
the technologies described here: URI
and Web Services. Together, they
move the WHO family of International
Classifications towards the so called
linked data approach.
URIs
A Uniform Resource Identifier (URI) is
a compact string of characters for
identifying an abstract or physical
resource. Even the URIs most known –
the subset called URL- are used as
addresses for networked resources
(web sites and pages, etc), the concept
behind allows to identify any kind of
resource, not necessarily corresponding
to something on the Web.
Web Services
The W3C defines a "Web service" as:
[...] a software system designed to
support interoperable machine-to-
machine interaction over a network,
usually based on SOAP and WSDL
standards. In the last years a category
of web services appeared that is based
on simpler representational state
transfer (REST) communications.
Linked data
The Linked Data approach exploits
URIs and, often, REST-based web
services to help interconnection of
structured data available on the Web,
leading towards the so-called Semantic
Web, and in particular towards the
specific concept of Semantic
Interoperability.
Tim Berners Lee provided four rules for
truly useful linked data:
1.Use URIs to identify things.
2.Use HTTP URIs so that these things
can be referred to and looked up
("dereferenced") by people and user
agents.
3.Provide useful information about the
thing when its URI is dereferenced,
using standard formats such as
RDF/XML.
4.Include links to other, related URIs in
the exposed data to improve discovery
of other related information on the
Web.
The URIs have been designed for the
ICD Foundation Component as well as
ICD-11 Linearizations and ICD-10.
Currently the services have been
deployed for the following content:
ICD Foundation Component
ICD11 Morbidity Linearization
ICD11 Mortality Linearization
ICD10 2010
ICD10 2008
Content negotiation for the format
The services behind the URIs provide
the classification in different formats:
The services support html, rdf/xml
and json-ld formats. To be able to
retrieve a specific format, we need to
use content negotiation by
appropriately setting the Accept
Header.
Content Negotiation for the
Language
The services want to be multilingual.
They support content negotiation using
Accept-Language header. Currently
only ICD-10 2008 has two languages
so this can be demonstrated only with
it.
Service URIs: Foundation
The ICD Foundation Component and
Releases of ICD are placed in different
URI paths.
Foundation URIs are as follows.
Top level
http://id.who.int/icd/entity
Returned Properties:
Title, Definition, Child
Individual Entity
http://id.who.int/icd/entity/{id}
Example:
http://id.who.int/icd/entity/1766440644
Returned Properties: Parent, Child,
Title, Definition, Long Definition,
Synonym, Narrower Term, Inclusion,
Exclusion, Body Site, Body System,
Causal Agents, Causal Mechanisms,
Signs And Symptoms, Genomic
Characteristics, Investigation Findings,
Type, Intent, Activity when Injured,
Object or Substance Producing Injury,
Mechanism of Injury, Place of
Occurrence, Substance Use
DIGITAL, MOBILE, NOW!
Scan this to get a digital version
ICD-11 Linearizations are available in
two variants, with or without minor
version.
URIs without minor version are as
follows:
Top level linearization:
http://id.who.int/release/11/{Linearization
Name}
Example:
http://id.who.int/icd/release/11/morbidity
Returned Properties: Title, Latest
Version, Version
Entity in a linearization:
http://id.who.int/release/11/{Linearization
Name}/{id}
Example:
http://id.who.int/icd/release/11/morbidity/2
1500692
Returned Properties: Title, Latest
Version, Version
URIs without minor version are as
follows:
Top level linearization:
http://id.who.int/release/11/{Minor
Version}/{Linearization Name}
Example:
http://id.who.int/icd/release/11/beta/morbi
dity
Returned Properties: Title, Definition,
Child
Entity in a linearization
http://id.who.int/release/11/{Minor
Version}/{Linearization Name}/{id}
Example:
http://id.who.int/icd/release/11/beta/
morbidity/1012371341
Returned Properties: Code, Parent,
Child, Title, Definition, Long Definition,
Inclusion, Exclusion, Index Terms,
Class Kind, Source
ICD-10 URIs
Similar URIs are available also for
ICD−10, e.g.:
http://id.who.int/release/10
http://id.who.int/icd/release/10/A00
http://id.who.int/icd/release/10/2010
http://id.who.int/icd/release/10/2010/A00
Conclusions
The availability of URI identifiers, and a
fully fledged API to access classification
entities, makes easier for any software
developer to invent new ways of
exploiting and combining WHO
classifications, thus opening them to a
wider and more intelligent use.
Further work is needed for securing
access to the API to registered users
only.
 WHO - FAMILY OF INTERNATIONAL CLASSIFICATIONS NETWORK ANNUAL MEETING 2013

12 – 18 October 2013
New Patient Safety Beijing, China
Indicators C319
Danielle Southern, Hude Quan
for the Quality & Safety TAG

INTRODUCTION Figure 1. Determination of possible PSIs

Health data coded using International
2,613 ICD10-CA
Classification of Diseases has been used to Type='2' Codes
identifyVerdana
the occurrence of adverse events that Title
36 Bold
may be linked to suboptimal safety and quality of exclusions
23 illegal uses of type '2'
care. Some countries code timing of diagnosis,
which is crucial to determine if the event Round 1

presented on admission or occurred after 2,590 ICD10-CA

Codes
admission. Using this unique data element, we
took two steps to produce novel patient safety Panel agreement
1493 rejected for PSI
Panel agreement

indicators (PSIs). 219 kept for PSI

METHODS & MATERIALS Panel disagreement

878
First, we queried 2,416,413 records in Canadian brought to Round 2
Discharge Abstract Database (DAD) for April 1, Title
for review

2009 through March 31, 2010. All listed

diagnosis codes were compiled, and 2,613 were Panel agreement
Panel agreement
141 rejected for PSI
found to have a type ‘2’ diagnosis designation, 285 kept for PSI

indicating that it was not present at time of

admission and thus arose after admission. Panel disagreement
but median score
Twenty three codes were excluded as coding suggests
Panel disagreement
299 rejected for PSI but median score
errors or inappropriate use of type ‘2’ diagnosis. suggests

Second, 7 Panelists were then asked to review all 153 kept for PSI

of the remaining 2,590 codes before meeting face

to face and to return their ratings (9-point scale).
Highest and lowest ratings were dropped and Figure 2. Frequency of PSI groupings by year
ratings were then analyzed to determine the
number of codes that had agreement as potential
PSI (all remaining panelists rated as 7 or higher)
or were rejected with agreement (all remaining
2005 4.53%
panelists rated as 1-6).

RESULTS
Of the 2,590 codes, 219 were agreed upon as
2006 4.74%
potential PSIs in round 1. Another 1,493
diagnosis codes were rejected with agreement.
The remaining 878 diagnosis codes produced
disagreements in panelist ratings, so these were
2007 5.00%
brought forward to the face-to-face meeting for
discussion. The detailed review and discussion of
these codes required 2 full days of panel
discussion. The second round of reviews
2008 5.37%
produced another 438 diagnosis codes for which
Conclusions
there was agreement that they were appropriate
as potential PSIs. In total, this 2-step process of
reviewing and rating type 2 diagnosis codes Title 2009 5.46%
produced a list of 640 codes that were
determined to be appropriate for consideration as
novel PSIs.
2010 5.54%
CONCLUSION
The methodological work presented here utilizes
the unique potential of diagnosis-timing indicators
to produce a clinically-relevant listing of diagnosis 2011 5.62%
codes that have potential as patient safety
indicators that may overcome some of the
notable shortcomings of existing patient safety 0.00% 2.00% 4.00% 6.00% 8.00% 10.00%
indicator systems. The resulting work has great
potential to inform future approaches to health
system monitoring and quality/safety
improvement internationally.
 WHO - FAMILY OF INTERNATIONAL CLASSIFICATIONS NETWORK ANNUAL MEETING 2013

12 – 18 October 2013
Determining the ICD Beijing, China
Shoreline C320

Authors: Jairus Quesnele, Molly Meri Robinson Nicol, Brian Basden,

Jeylon Buyukdura, T. Bedirhan Ustun
World Health Organization, Geneva, Switzerland
Abstract The ICD Shoreline is the conceptual boundary between the concepts that will exist as precoordinated stem codes
and those that will exist as post-coordinated entities. Drawing this shoreline specifies what detail appears in the Mortality
and the Morbidity Linearizations, and how. This process requires adopting certain classification principles and addressing
issues that will be highlighted. Addressing these will help produce a robust classification system.

Introduction Title where post-

Establishing a concept-based boundary
between entities that are
precoordinated and those that will be
post-coordinated is a critical
component of ICD Revision. In the
shoreline analogy, precoordination
refers to land and post-coordination
refers to the sea. Precoordinated
concepts are fully specified stem
codes. Post-coordinated concepts are a
systematic combination of stem codes
extended with sanctioned extension
code entities to add detail.
Determining which ICD entities should
be precoordinated and which post-
coordinated will establish the Mortality
linearization as well as the substance
of the Morbidity linearization. There
are many principles which govern this
process. The purpose of this poster is
to highlight the principles and key
issues that are essential to the
development of the revised ICD as a
robust, fully-comprehensive
classification system.
Precoord.
Post-coord.
General Concepts
1. Certain ICD concepts will be fully
specified in their detail - this is called
precoordination, and all precoordinate
concepts are "stem codes".
2. Post-coordinating a concept is
accomplished through adding one or
more allowed extension codes to an
established precoordinated stem code.
3. Post-coordination is not permitted
in the Mortality linearization, therefore
all concepts required for Mortality will
be precoordinated.
4. All precoordinated Mortality
concepts will also be precoordinated
stem codes in the Morbidity
Linearization.
5. For all shoreline decisions, the
following rules are considered:
a) Legacy
b) Scientific evidence
c) Taxonomical and ontological rules
for consistency
d) Practicality of coding
e) Frequency in practice setting
f) Utility (Clinical and Public health)
Key issues
1. When generic dimensions exist
across chapters they will be
contained in the X-Chapter.
Precoordinated entities that include
this generic detail should instead be
post-coordinated. Post-coordination
with the R-Chapter works in a
similar fashion for clinical forms and
findings. Entities that are
inappropriate
Title for the Morbidity
Linearization and cannot be post-
coordinated will appear only in the
Foundation Component and possibly
in a Specialty linearization.
2. When a category is secondarily
parented elsewhere, the order of
the children within the category is
presently grandfathered from the
linearization parent.
3. Linearization parenthood should
depend first upon etiology, if it may
be known. The two chapters of
Infectious Diseases and Neoplasms
have been given primacy in the
Morbidity linearization; all their
children should be linearized there.
4. There is presently underutilization
of multiple parenting. Many children
that clearly relate to two or more
topic areas may not yet have their
multiple parents specified.
5. Groupings should be used in a
balanced manner. Having either
excessive layers of groups before
children or too few groups make
the classification system inefficient.
Architectural-related issues
Title
1. Additional detail not relevant for the
Morbidity linearization should be
expressed in the relevant Specialty
linearization. For every detailed
Specialty linearization child, there
will be a parent/ancestor that exists
in the Morbidity linearization.
2. There will not be any items in
Mortality that are not expressed in
the Morbidity linearization.
3. If a proposed new entity is required
in the Mortality linearization or in
the Morbidity linearization at the 3rd
or 4th depth level or lower, then the
essential parameters in the content
model should be populated with
content.
4. Under certain parents
coordination is not possible and long
heterogeneous lists of children
exist, these lists should be more
practically grouped.
5. Residual codes in the revised ICD
will be automatically generated, and
will include “Other” entities coded
with a ‘Y’ and “Unspecified” entities
coded with a ‘Z’, similar to the
legacy ".8 and .9" codes from ICD-
10.
6. Exclusion terms must be expressed
in the relevant linearizations as
codable entities, only. Therefore
each exclusion must be included as
a foundation entity.
7. To retain a consistent architecture,
chapters that contain entities which
apply throughout ICD (i.e.
infections) will contain an alternate,
high-level grouping structure that is
organized by body system, similar
to the organizational structure of
the rest of the classification.
8. Entities that appear grey in the ICD
Beta Browser indicate multiply-
parented entities that are linearized
elsewhere.
Naming convention issues
1. Entities that include ‘classified
elsewhere’ will be deprecated.
2. Entities that contain the words
‘Specific’ or ‘Certain’ in the title should
have corresponding children, and will
be reviewed for Conclusions
possible
reorganization.
3. There should be no acronyms or
abbreviations in ICD entity titles.
Desired alternate terms should be
included as a synonyms.
4. The use of disease, disorder, and
syndrome must be consistent with the
WHO definition of these entities.
5. WHO has created a feature on the
ICD Beta browser, under the ‘More’ tab
called ‘Errors’ (one must be logged in).
Here, a list of duplicate terms and
various linearization errors are
identified, highlighting priority
corrections to be made.
 WHO - FAMILY OF INTERNATIONAL CLASSIFICATIONS NETWORK ANNUAL MEETING 2013
Authors: Jairus Quesnele, Molly Meri Robinson Nicol, Brian Basden,
T. Bedirhan Ustun
World Health Organization, Geneva, Switzerland
Abstract ICD foundation entities will be presented and organized in several different, overlapping lists called
linearizations. While the Mortality and Morbidity linearizations, as well as some others, will be set by WHO, other
linearizations offer flexibility and may address specific purposes. As there is some confusion as to how each linearization is
developed and what its purposes will be, this report will further explain how ICD linearizations may be created and used.
Introduction
ICD foundation entities will be
presented and organized into several
different, overlapping lists called
linearizations. A primary goal of ICD
Revision is to address multiple use-
cases with a coherent, fully-
comprehensive classification.
Currently, the Mortality and Morbidity
linearizations are the priority
linearizations. Other linearizations may
include those for High and Low
Resource Primary Care Settings,
National and other Specialty
Linearizations, Research, or Verbal
Autopsy. Given these myriad options, a
clear description of the purpose of
each linearization may be useful.
Mortality Linearization
In the Mortality linearization, included
entities should be relevant causes of
death. Post-coordination is not used in
Mortality, therefore the entire Mortality
linearization consists of precoordinated
concepts. ICD-10 legacy has strong
influence when debating inclusion of a
Mortality item in ICD. Reporting
frequency, both national and
international, is also an important
consideration in the inclusion of an
entity in the ICD Mortality
linearization.
The Mortality linearization is a proper
subset of the Morbidity Linearization.
All entities included in the Mortality
linearization will also be included as
precoordinated concepts in the
Morbidity linearization where, as stem
codes, they may be expanded with
additional detail taken from the post-
coordination space.
There is a telescoping principle within
the various ICD linearizations, with
particular focus on the Mortality and
Morbidity linearizations. The Mortality
linearization codes will essentially be
added to, ‘zooming in’ or condensed,
‘zooming out’. This principle is merely
a guide, rather than a hard rule.
(Figure 1: red arrow).
What’s in an ICD
Linearization?
Morbidity
MorbidityLinearization
Linearization
The ICD Morbidity linearization is the
primary international reference for
reporting and data exchange. The
Morbidity linearization will be fully-
comprehensive, while avoiding
unnecessary explosion of the coding
space. Additional detail in the
Morbidity linearization will be
expressed through extension codes,
but mayTitle
still remain precoordinated in
the ICD Foundation to be expressed in
other linearizations. In keeping with
the telescoping principle, additional
linearizations may be created as an
extensions of the Morbidity
linearization. This permits other
linearizations to include more detailed
stem codes, if desired.
Mort/PCHigh
PC – Low 1 Mort/PCHigh
Mort/PCHigh
PC – Low 2
Mort/PCHigh
Mort/PCHigh 22
PC – Low 3
Mort/PCHigh
Mort/PCHigh
Mort/PCHigh
Mort/PCHigh
Mort/PCHigh
PRIMARY CARE Low PRIMARY CARE High Resource MORBIDITY
Resource
(Verbal Autopsy ?) MORTALITY
Figure 1: Linearization telescoping principle
Specialty Linearizations
Specialty care may require more detail
than the standard Morbidity
linearization. Specialty linearizations
allow for increased expressivity to
accommodate this need. Identified
specialties will have the liberty to
further develop these linearizations
with the support of WHO. All additional
entities in the Specialty linearizations
will have identifiers to enable tracking
for data comparability.
Research Linearizations
The Research linearization also
accommodates expression of further
detail, including additional items from
the Foundation. This linearization may
exist outside of the telescoping
principle, reorganizing ICD entities in a
different way. The correspondence of
these versions for equivalence
purposes is accomplished through the
Foundation.
12 – 18 October 2013
Beijing, China
C321
Title
National Linearizations
For national purposes, countries that
require additional detail when adopting
ICD may create National
Linearizations. These will be similar to
the Clinical Modifications generated
previously. The new mechanism,
however, will engender better
coordination mechanisms to avoid non-
standard development. This will
require that countries add all additional
items to the Foundation and identify
the linearization parents and shoreline
(either precoordination or post-
coordination). WHO will provide the
infrastructure and the tools. In this
way, it is envisaged that comparability
between different national
linearizations will be maintained
through the Foundation.
Morbidity111
11
Morbidity112
12 Morbidity121
Morbidity131
13
Morbidity132 Primary Care Linearizations
Morbidity133
21 Morbidity211
WHO will generate specific
Morbidity221
linearizations for both high and low
resource settings in Primary Care (PC).
Morbidity222
Morbidity311
These linearizations will be condensed
31
Morbidity312
from the Mortality and Morbidity
32
Morbidity321
linearizations, and include the most
33
relevant entities in primary care. The
Morbidity341
Low Resource PC linearization will be a
34
Morbidity342 condensed version of the High
35 Morbidity351 Resource PC linearization.
Extensions
Verbal Autopsy Linearization
International National Linearizations
Specialty - Research
In certain low resource settings, verbal
dictation of causes deaths is utilized, a
process known as Verbal Autopsy. To
Conclusions
support reporting in this manner, WHO
will provide a linearization which
mirrors the Low Resource PC
linearization while supporting this use-
case.
Sanctioning Tables
Post-coordination mechanisms may
apply to all linearizations other than
the Mortality linearization, and will be
widely implemented in the morbidity
linearization. To avoid double coding
within these linearizations, sanctioning
tables are being created. These tables
will identify the combination of a stem
codes and relevant extension codes as:
• Applicable,
• Required, or
• Non-applicable.
 WHO - FAMILY OF INTERNATIONAL CLASSIFICATIONS NETWORK ANNUAL MEETING 2013

Ensuring a Seamless Transition: 12 – 18 October 2013

Beijing, China
Results of the Morbidity Stability Analysis
C322
Authors: Molly Meri Robinson Nicol, Linda Best, Maren Hopfe
World Health Organization, Geneva, Switzerland

Abstract Several key stakeholders to ICD Revision have expressed concern regarding the potential for disruption to existing health
information systems, DRG systems, data collection systems, or other environments into which ICD-10 or a national clinical modification of
ICD-10 has been integrated. This poster describes the methodology of the work undertaken to complete a stability analysis of the
revised ICD and the existing ICD-10, with the addition of the national clinical modifications thereof, specifically with a focus on the
Morbidity Linearization.
Title
These efforts ensure that each entity, both those from the selected national clinical modifications and from the international version, can be
accounted for within the revised ICD. This work is done to prepare for, and facilitate, a seamless transition at the time of ICD publication
by ensuring continuity in data collection between ICD-10 and the revised ICD

Introduction Results
Several countries, such as Australia, After finalizing the manual check the following mapping results emerged (table 2, figures 1-5).
Canada, Germany and the United
States of America, have extended
and modified ICD in order to address
specific country requirements. Given
that these modifications may be the
basis for existing national health
information systems, data collection, Title
reimbursement, and health policies,
the current revision process raises
concerns regarding transition to the
revised ICD.

Main objectives:

• identification of national Table 2: Mapping results of the country specific codes to the revised ICD * Codes identified as AM-specific,
some may be reconsidered as
modification specific codes
inclusions
• identification of all ICD-10 codes ** After completing the CA Stability
relevant for the DRG systems Analysis, we were advised that the file
• identification of the status of the used included outdated information.
ICD-10 codes within the ongoing Updated files were received 26 August
2013, and efforts to update the
revision process, and analysis are underway.
• mapping of the ICD-10 codes to
the revised ICD. Results show that 78.9% of codes from ICD-10 (WHO)
remain acceptably constant in the revised ICD with a
The goal is to ensure that each similar percentage for ICD-10-GM due to the inclusion
individual ICD-10 code (country of many GM-specific codes in the start-up list for ICD
specific and WHO) is identified and Revision. Although the total number of entities
accounted for in the revised ICD. remaining constant from ICD-10-CM is similar to that
of ICD-10-GM, the percentage is much lower due to
the greater number of additional codes added to cover
Methods & Materials the vast expansion of detail of laterality, severity, and
other dimensions in ICD-10-CM. ICD-10-AM, 7th
The source files used for the analyses are listed in table 1. The status (see edition shows similar results. ICD-10-CA also indicates
below) of each country specific code (XM codes) as related to the revised ICD was good accord, but as noted in the table above, after
identified using the Vlookup function in Excel 2010 followed by manual verification completing the CA Stability Analysis, we were advised
of matched and unmatched codes. that the file used included outdated information.
• XM Codes already included in the Foundation unchanged (code match & title match) – Updated files were received 26 August 2013, and
automatic mapping efforts to update the analysis are underway.
• XM Codes which have a code match but title mismatch in the Foundation – manual mapping
• XM Codes which have a title match but code mismatch in the Foundation – automatic map, Conclusions
Conclusions
manually verified
• XM Codes which have a title match but with a different code match – automatic map, manually The concerns raised by stakeholders are valid and do
verified require investigation. The methodology devised for this
• XM Codes which have a title match but no code at all in the Foundation – automatic map, investigation was termed Stability Analysis.
manually verified
• XM Codes which have neither a code match nor a title match in the Foundation – manual
Although not yet completed, the interim results of
mapping
stability analysis presented at this time provide a
generally positive view, supporting the idea that the
transition to the revised ICD in countries will be
reasonably feasible with minimal disruption to existing
health systems from a purely coding-based perspective,
political implications aside, for Germany and Australia.
Accommodations for the United States modification can
be made through judicious expansion of the ICD
foundation paired with development of mappings which
utilize the post-coordination dimensions of the ICD X-
chapter, particularly severity and laterality.

Future work will include update the analyses after the

Table 1: List of source files
InEK: German Institute for Hospital Remuneration System
NCCC: National Casemix and Classification Centre (NCCC), University of Wollongong (Australia)
CDC – NCHS: Center for Disease Control and Prevention – National Center for Health Statistics (USA)
DIMDI: German Institute of Medical Documentation and Information
WHO: World Health Organization
Morbidity linearization shoreline is stabilized, a process
now underway. It will also be necessary to work with
the owners of each clinical modification, generally the
relevant WHO Collaborating Centre, to verify the results
of these analyses and to prepare the strategies
necessary to move forward with a seamless transition
from ICD-10 to the revised ICD.
 WHO - FAMILY OF INTERNATIONAL CLASSIFICATIONS NETWORK ANNUAL MEETING 2013
Kostanjsek N1 on behalf of WHO CTS, RSG and the
1CTS,
Abstract To ensure that ICD-11 is working well, it needs to be systematically field tested in different settings, across the
world. Only through exposure and testing in the real world we can assess the “fitness of ICD-11” and make the necessary
enhancements as needed. This poster provides an overview of the envisaged Field Trial objectives, core study protocols
and management arrangements.
Objectives & key assessments
The overarching objective of the Field
Trials is to ensure systematical testing
of ICD-11 before its use to increase
consistency, identify improvement
paths, and reduce errors.
The basic aims include:
• the testing the “fitness of ICD-11 for
multiple purposes”;
• and ensuring the comparability
between ICD-10 and ICD-11.
To achieve the overall objective and
specific aims the Field Trials will focus
on the following key assessments
and questions:
Applicability (Feasibility): Is the
classification easy to use in real-life
contexts and settings? How easy can
first time ICD-11 user accomplish their
documentation or coding tasks? Once
familiarized with ICD-11, how quickly
can users perform their tasks? How
can the use of ICD-11 made a pleasant
experience?
Reliability (Consistency) Is the
classification rendering the same
results when applied by different
users? Do two or more different users
code the same diagnosis with the same
ICD code? What are the sources of
discrepancy? What are the factors to
improve comparability and
consistency?
Utility (Benefits) Is the classification
providing useful information and
adding value for guiding diagnosis and
enhancing data capture. Does the
classification enable better
documentation, aggregation,
comparability and re-use of data? Does
it allow better resource allocation?
12 – 18 October 2013
ICD-11 Field Trials Beijing, China
C323
WHO FIC Network
World Health Organization, Geneva, Switzerland
Morbidity
Field Trial
Linearization
Studies Study 2: BridgeTitle
Coding
The Field Trials will comprise a series The study will assess
of (mandatory) core studies and a the level of
range of (optional) additional studies. agreement between
coders when coding
Core studies the same diagnosis
The core studies include the following using ICD-10 and
three study protocols: ICD-11 in an effort to ascertain and
enhance the comparability between
Study 1: Reliability and Feasibility
ICD-10 and ICD-11.
The study aim is to test the reliability
Title of ICD-11 in various
and feasibility
Study 3: Basic
settings, formats and versions; aiming
Questions
to maximize consistency and limit
Through a Key
errors.
Informant Survey and a
Consensus Conference
this study is seeking to
identify the spectrum of
Photos: Mock-up interviews conducted in Chinese and Korean Hospitals in pre-
paration for the implementation of the Field Trial Pilot Phase for ICD-11 Ch. 23. opinions as well as a
The reliability will be tested through consensus statement
on conceptual as well as operational
evaluation and comparison of inter-
issues related to ICD-11.
rater reliability. Case Summaries, Live
or Video Cases will be coded by more The Basic Question will address topics
than one rater and a comparison will such as:
be made between multiple raters • needs and usage;
recording of the corresponding ICD-11 • diseases conceptions & delineation;
code. The study will also identify the • coverage & structure;
reasons for discrepancy between rater • terminology principles;
assigned codes, in an effort to improve • coding;
definitions, coding instructions, tools
• ICD-11 and other WHO FIC.
and criteria.
The feasibility of ICD-11, will be Additional studies
studied through a series of questions In addition to the core studies
which raters will be ask to respond to additional study protocols on specific
after having completed the reliability issues may be developed for optional
testing. implementation.
Study 1: Instruments & procedure
Field Trial Structure
The field trial structure will be two
tiered, the first tier at national level
consisting of Field Trial Centers (FTCs).
A Field Trial Centre is a WHO approved
study centre that will manage the
implementation of Field Trials at
country level and report directly to
WHO. WHO Collaborating Centers,
ICD-11 Topic Advisory Groups or other
organization with sufficient
implementation
. capacity can serve as
FTC.
The second tier is consisting of multiple
Field Trial Sites (FTS) coordinated and
supervised by a FTC. FTS should
represent different settings (e.g
primary care, general health care,
research settings).