DESIGN AND EVOLUTION OF HOME-CARE WORKFLOWS

USING ONTOLOGIES AND MODEL TRANSFORMATIONS

Sabrina Zefouni

1,2

, Elyes Lamine

1,2

, Rémi Bastide

and Hervé Pingaud

ISIS - CUFR J.F. Champollion, Campus Universitaire, Rue Firmin Oulès, Castres, France

EMAC, Centre de Génie Industriel, Jarlard, Route de Teillet, Albi, France

Keywords: Workflow, Ontologies, Homecare, Coordination.

Abstract: This paper investigates the use of workflows to automate the delivery, coordination and monitoring of

home-care processes. Workflow models of these processes need to be customized, resilient to change,

collaborative and also need to account for the temporal aspects of care (treatment duration, frequency of a

medical acts, etc.). To this end, an approach based on ontologies (home-care ontology and BPMN

ontology), rules and patient’s profile is proposed for the construction of workflows supporting the

coordination and continuity of home-care processes.

1 INTRODUCTION

Home-care presents two majors problems, namely

coordination and continuity of care. These problems

have been highlighted by the document

MIN/DHOS-/DSS-CNAMT/2002/n°610, and

several others (Arundel et al, 2001) (Bricon-Souf et

al., 2005) and (Woodward et al., 2001). From our

studies in the home-care area, these drawbacks are

still valid. 1) Lack of inter-organizational

communication or coordination of care: there are

several associations, organizations, stakeholders that

manage employees to work at the patient's home.

However, there is no well-established means of

communication between these different stakeholders

(Arundel et al., 2001) and (Helleso et al., 2005).

Indeed, the only means used by many home-care

organizations to enable this communication, is the

so-called "liaison logbook" i.e. a physical notebook,

usually unstructured and kept at the patient's home

(Bastide et al., 2010) and (Lamine et al., 2010). 2)

Problem of care continuity: in the literature, there

are a multitude of definitions for continuity of care

like (Woodward et al., 2004) and (Shortell, 1976). In

homecare, the need for continuity stands at two

levels: the intervention level where the stakeholders

must ensure continuity and consistency of care

provided. But also, over time, during different steps

of patient treatment: moving from one institution to

another.

To ensure the coordination and continuity of

home-care, workflow management systems appear

to be appropriate. WfMC (Workflow Management

Coalition) (Hollingsworth, 1995) has defined

workflow. The role of a workflow management

system, in home-care, is to coordinate the work

involved by transmitting the necessary information

and tasks, with different stakeholders, depending on

the condition of the patient's intervention plan and

the role of the interveners. For example, the nurses’

day-to-day tasks vary according to the patients’

intervention plan, and the storage and exchange of

the relevant information would improve the

continuity of care.

However, the specific characteristics of

processes involved in homecare make the design of

their workflow a challenge. The use of ontology to

help in the design and evolution of these workflows

is proposed in this paper

In section 2 we discuss the need for workflow in

automating the home-care processes and we point

out the challenges for this. Section 3 is focused on

the presentation of proposal approach to design a

workflow for home-care. We discuss the related

work and our objectives on the section 4. Finally, we

conclude with a summary and outlook.

2 PROCESSES IN HOME-CARE

The activity of home-care gives rise to

communication and coordination problems. To

ensure the coordination and continuity of care, and

to improve its quality, the caregivers involved on

231

Zefouni S., Lamine E., Bastide R. and Pingaud H..

DESIGN AND EVOLUTION OF HOME-CARE WORKFLOWS USING ONTOLOGIES AND MODEL TRANSFORMATIONS.

DOI: 10.5220/0003779302310236

In Proceedings of the International Conference on Health Informatics (HEALTHINF-2012), pages 231-236

ISBN: 978-989-8425-88-1

 2012 SCITEPRESS (Science and Technology Publications, Lda.)

home-care are mandated to: 1) ensure the accuracy

of information, by automating the transmission of

information and by avoiding several transcriptions

for the same information; 2) Guarantee the timely

transmission of information to the appropriate

stakeholders, be it information about the patient's

condition or other logistical information; 3) Ensure

the transmission of information about the actual

medical or paramedical interventions performed by

the caregivers; 4) Ensure the storage of information

about the evolution of patient's health conditions.

2.1 Home-care Processes

In the literature, there are many research works

about home-care processes and particularly on

medical home-care (Jebalia, 2008), (Ben Bachouch

et al., 2009), (Lunn et al., 2003), (Hamek, 2005) and

(FNEHAD, 2009). In (FNEHAD 2009) the

"National federation of institutions hospitalization at

home" (FNEHAD) provides a mapping process for

patients under medical home-care, considering that

computer support is still underdeveloped in

institutions of medical homecare. This work aims to

structure information systems for home-care and

stands as a reference in this area, as it encompasses

all processes presented in other works. Figure 1

shows the main processes involved in home-care,

following ISO 9000: 2000 (ISO, 2000) which

classifies processes into three types: management,

operational (management of patient) and support

(Human resources, purchasing and logistics...).

Figure 1: Home-care mapping processes.

2.2 Workflow in Home-care

To cope with the drawbacks addressed above, we

suggest the automation of some processes of patient

management. In this paper, we focus on the delivery,

coordination and monitoring of care.

We distinguish two sets of processes: The first

set includes the "admission request",

"preadmission", "billing stays" and "discharge"

processes, which are prone to be automated by a

classical workflow management system. The second

set of processes includes "delivery of care" and

"coordination and monitoring of care" that are the

main topic of the present paper. The characteristics

of these processes make the design of workflows for

their automation a difficult task. This set of

processes has the following characteristics: 1) High

degree of customization (Dadam et al., 2000),

(Dazzil et al., 1997) and (Quaglinia et al., 2001):

each patient is a specific case due to particular health

conditions, social networking, geographic location,

etc. 2) Collaborative nature: (Arbaoui, 2008) and

(Cunin, 2000). The homecare processes present a

strong human component. Such a process is usually

long-running, distributed among several actors with

various levels of autonomy, and subject to dynamic

change (Liu et al., 2008). This collaborative process

is made of the assembly of distributed business

processes in partner organizations. Its behaviour is

related to the effective contributions of partners,

implemented jointly to achieve a common goal. 3)

Dynamicity: (Dadam et al., 2000) according to the

changing health of the patient and his environment.

In general, the process it required to quickly and

easily adapt to a changing environment. 4) Temporal

aspects: (Dadam et al., 2000) the tasks in homecare

have frequency and duration which are inconvenient

to represent in conventional workflow models.

3 PROPOSITION

3.1 Context

A workflow to deliver, monitor and coordinate

home-care is a sequence of interventions, as shown

in figure 2. This workflow depends on the

intervention plan defined by the professionals that

manage home-care institutions, and depends on the

profile of patient: diseases of patient, the presence of

family caregiver, etc.

Figure 2 shows - using BPMN (Business Process

Management Notation) - the sequence of

interventions performed within a two-days’

timeframe for a specific patient (Mrs Bissière that

has peritonitis vesicular, the case study detailed on

section 4), with the distribution of interventions

amongst stakeholders (nurse, physiotherapist, etc.).

The intervention plan defines the list of provisional

care to be performed, including actions and

treatment, prescribed or not, as well as the frequency

and schedules, and specify the appropriate

takeholders (nurse, cleaning agent, etc.) to perform

each actions.

HEALTHINF 2012 - International Conference on Health Informatics

232

Figure 2 shows one pool representing the

patient's home, containing four lines representing

four different categories of actors involved in the

home-care of a specific patient: Nurse, Cleaning

agent, care aid and Physiotherapist. The first day

features five interventions: In the morning, the

Nurse has to perform perfusion, alimentation via

parenteral route, vacuum assisted closure,

preparation of pill box and blood drawing. In the

evening, she performs perfusion and alimentation

via parenteral route. Cleaning agent performs

cleaning room, Care aid performs help with dressing

and bathing and changes the cover, and

Physiotherapist performs physiotherapy. Second day

features the same five interventions but a difference

appears at tasks level of nurse in the morning, when

she performs only three tasks that are perfusion,

alimentation via parenteral route and vacuum

assisted closure. Each of these interventions is done

at a specific moment, e.g. the intervention of

cleaning agent must be performed in the morning

and be followed by the intervention of nurse.

Figure 2: Home-care Workflow.

An intervention is a sequence of three actions,

namely: 1) consult transmissions: consist in

consulting eventual transmissions (remarks, alert,

etc.) noted by a precedent stakeholder. 2) Inform on

performed acts: display the acts to be performed

during this intervention. It is the dynamic and

adaptable part of the intervention. It changes each

day, and for each stakeholder. 3) Transmit

information and alerts to other stakeholders: consist

in writing information about patient to inform or

alert others stakeholders about any change on

patient’s state or environment.

3.2 Proposed Approach to Design

Home-care Workflows

In recent years the concept of ontology has been

largely studied in computer science. Ontologies

allow for the sharing of knowledge to solve

problems of semantic heterogeneity. There are

several definitions of the concept ontology amongst

which we choose "ontology is a formal, explicit

specification of a shared conceptualization"

(Thomas, 1993).

The works we mostly refer to is that of (Valls et

al, 2001). The authors propose ontologies for

medical home-care. The design of these ontologies

is based on a model of home healthcare defined by a

European consortium of professional homecare,

European project K4Care.

Figure 3 shows the position of our proposal

within the reference architecture of workflow

management systems proposed by the WfMC

(Hollingsworth, 1995). This architecture consists of

five interfaces with the following components: 1)

process definition tools: tools for graphical

modelling of the process to automate and deploy.2)

Workflow client applications: any applications to

communicate to-do lists, messages, etc. to actors.

There are applications that interact with the user. 3)

Invoked applications: any application: DBMS, web

service, etc. called or invoked during the deployment

process automatically without intervention or

interaction of a user. 4) Other workflow enactment

services: any other workflow management system

able to communicate with the workflow

management system. This ensures the

interoperability between different workflows

management systems on the market. 5)

Administrative and monitoring tools: provides

access to admin tasks, such as suspension of a task,

stopping a process, etc.

Based on this reference architecture for

workflow management systems, we propose the use

of ontologies at level 1 and 3: 1) in the process

definition tool, the goal is to help the construction of

patient personalized care workflow. 2) At the

invoked applications, the goal is to guaranty

interoperability and flexibility of workflow at run

time.

Figure 3: Position of our proposition in the reference

architecture for workflows systems.

In this paper, we discuss workflow design. The

idea is to use ontologies for help designing the

workflow models for delivery, monitoring and

DESIGN AND EVOLUTION OF HOME-CARE WORKFLOWS USING ONTOLOGIES AND MODEL

TRANSFORMATIONS

233

coordination of homecare. The definition of an

intervention plan maps to the definition of a

workflow model. They both define the distribution

of tasks (care intervention) to actors (homecare

stakeholders) in order to achieve the specific goal to

improve the quality of home-care. But an

intervention plan defines frequency of intervention

and their duration over the treatment period.

Transformation Process. Figure 4 shows the steps

to design customized BPMN process models for

each patient. The resulting process models will be

executed on workflow system.

We use these components: 1) a knowledge base

is a fusion of the home-care ontology (presented in

the next sub-section) and the BPMN ontology

(Ghidini et al., 2011). This knowledge base contains

patients’ profiles and rules about actors and actions

involved in treatment of a specific disease. 2) A Java

application is based on OWLAPI. OWLAPI is a java

API and reference implementation for creating,

manipulating and serializing OWL Ontologies. This

application queries the knowledge base about an

intervention plan of specific patient. The output of

this application is an XML file. The XML file

defines an intervention plan of specific patient as a

BPMN process model. 3) Tools are used for model-

transformation. The XML file resulted in the last

step must be transformed in a format suitable to be

interpreted by a workflow engine. Tools such as

ATL or XSLT are well suited, and allow

transforming the XML file into a standard BPMN

2.0 file.

Figure 4: Methodology for customized workflow design.

Embryo of Ontology for Home-care. Little

ontologies have been built in the home-care area, we

present here two projects: ABAH: An Agent-Based

Architecture for a Cooperative Information System

(Zarour et al., 2007), an Algerian project with the

goal to ensure interoperability in a multi agent

system. K4CARE European project (presented

above) with their two ontologies APO: Actor Profile

Ontology and CPO: Case Profile Ontology: CPO is

for representing symptoms, diseases, syndromes, etc.

Based on these ontologies, and our own surveys

of current practices in homecare, we propose a high

level ontology of home-care. It is represented as the

usual RDF triples where circle represent class

(concepts of home-care domain) connected by

relationship arrows (figure 5). For ease of

reading we have not presented the inverse functions

(relationship)

This ontology contains global information about

a structure and organization of home-care (actors,

roles, etc.) as well as profile of patients and the task

and actors needed for each disease.

To build this ontology of the home-care domain,

we followed a synthesis of different methodologies

in literature such as the ones proposed in (Uschold

and King, 1995), (Staab, 2001). The ontological

engineering process for building the home-care

ontology has been performed under four steps: 1)

Step 1: this step consists in identifying the goal to

define the ontology, the users of this ontology and

scenarios of using this ontology.

Figure 5: Main classes and relations of the ontology for

home-care.

2) Step 2: based on the result of the first step, we

define the basics concepts and relationships for this

ontology. 3) Step 3: we choose a tool to design the

ontology and start buiding frome the general

concepts to the more specific. 4) Step 4 is an

iterative step, consisting in maintiaining the the

consistency of the ontology.

4 CASE STUDY

The proposition consists in indicating the profile of

the patient: pathology, family caregiver, etc. The

application, based on the ontology and the patient

profile, proposes an intervention plan: set of actors

involved in home-care, tasks, sequence of

interventions and frequency of these interventions.

The solution is based on populating the ontology by

information of patient.

The case study concerns Mrs Bissière. She is 72

years old and has a Peritonitis Vesicular.

In the knowledge base we have this information.

Data that the professional introduces on the

Knowledge base

 MrsBissière hasDisease

PeritonitisVesicular

HEALTHINF 2012 - International Conference on Health Informatics

234

Data that models general domain knowledge:

 PeritonitisVesicular InvolvesActor:

Nurse, CleaningAgent, CareAid,and

PhysicalTherapist

We distinguish between two kinds of rules

(expressed in SWRL): 1) rules that relate to concepts

of home-care ontologies. 2) Rules that make a

mapping between the two ontologies (home-care

ontology and BPMN ontology).

These two types of rules are illustrated by the

example below: the first rule indicates that, if patient

X has a disease Y and disease Y involves the

intervention of actor Z, then we deduct actor Z. The

second rule indicates that, if there is an actor X then

there is a lane X (actor that is a concept from the

home-care ontology, mapped to the equivalent

concept of lane in the BPMN ontology).

 Patient(?x),hasDisease(?x,?y),InvolvesAct

or(?y,?z)->Actor(?z)

 Actor (?x)-> Lane (?x)

We request the knowledge base to include the

BPMN concepts to generate the home-care

processes. An example of query is presented below

Query

: lane and (Actor and

isInvolvedToCareDisease some (Disease and

(isDiseaseOfPatient value MrsBissière)))

Results: Nurse, CleaningAgent,CareAid and

PhysicalTherapist

In this case study, the knowledge base includes the

details of the patient profile: Mrs Bissière has

vesicular peritonitis. Then, we query the knowledge

base about the concepts of BPMN (lane) involved in

the process of home-care of Mrs Bissière. It is

possible to answer to this query based on knowledge

and rules contained in the knowledge base.

5 RELATED WORK

There are several studies on the use of workflow in

health-care, especially at the hospital. In the study of

(Dadam et al., 2000), the authors discuss the

challenges to elaborate workflow on clinical

domains. They point out the critical role of

workflows in the clinical area. Like Song and al

(Song et al, 2006) where the authors note that

improving healthcare workflows is very important

for improving healthcare quality and efficiency. The

authors of (Quaglinia et al., 2001) propose to design

of a flexible workflow for the management of a

stroke at hospitalization. This work is focused on a

specific problem (stroke case) whereas in the case of

home-care, each patient may suffer from various

diseases including stroke. In addition, the issues of

home-care differ from those encountered in the care

in the hospital, where all stakeholders work within

the same organization and use the same information

system, with the ability to see to discuss the status of

patients.

Many authors use ontologies combined with

workflow in order to enhance flexibility or agility. In

(Bouzguenda et al., 2008) the authors are interested

to a collaborative environment to implement the

WIO (Workflow Inter-Organizational). The authors

propose a solution based on multi-agents systems for

the execution of flexible web services to facilitate

partner search and ontologies for searching and

selecting the partners. In (Vieira et al., 2004) the

authors suggest use of ontologies to achieve flexible

execution of workflow. In our work, we are

interested in the design of customized workflow

models for each home-care patient. In our

knowledge there are no others works with the same

goals.

6 CONCLUSIONS

In order to improve communication and continuity

in home-care, we propose system architecture based

on a workflow engine to distribute tasks among

home-care stakeholders. However, the

characteristics of home-care processes make the

design and enactment of home-care workflow a

challenge. The goal of this architecture is to respect

the reference architecture for workflow management

systems while relying on ontologies to overcome

these identified obstacles. We propose an approach

using two ontologies (home-care ontology and

BPMN ontology) to help designing home-care

workflows. Our on-going work consists in

completing the transformation pipeline in order to

generate process model suited to be interpreted by a

workflow engine.

REFERENCES

Arundel C., Glouberman S., Kun L. 2001. Telehealth and

the global health network. In the 21st century, From

home care to public health informatics in Computer

Substudy, vol. 15.

Bricon-Souf N., Anceaux F., Bennani N., Dufresne E.,

Watbled L., 2005. A distributed coordination platform

for home care: analysis, framework and prototype.

International Journal of Medical Informatics.

Woodward C., Abelson J., Hutchison B., 2005. My home

is not my home anymore: improving continuity of care

DESIGN AND EVOLUTION OF HOME-CARE WORKFLOWS USING ONTOLOGIES AND MODEL

TRANSFORMATIONS

235

in homecare. The Canadian Health Services Research

Foundation web site, (online: www.chrsf.ca).

Helleso R., Sorensen L., Lorensena M., 2005. Nurses’

information management across complex health care

organizations. International Journal of Medical

Informatics

Bastide R., Zefouni S., Lamine E., 2010. The Homecare

Digital Ecosystem: an Information System Support

Architecture. In: IFIP Conference on Personal

Wireless Communications, Knowledge Village, Dubai,

UAE, IEEE

Lamine E., Zefouni S., Bastide R., 2010. A system

architecture supporting the agile coordination of

homecare services. In : IFIP Working Conference on

Virtual Enterprises (PRO-VE), St-Etienne, 2010, Vol.

336, Springer

Woodward C. A, Abelson J., Tedford S., Hutchison B.,

2004. What is important to continuity in home care?.

Perspectives of key stakeholders Social Science &

Medicine 58 177–192

Shortell, S., 1976. Continuity of Medical Care:

Conceptionalization and measurement. Medical Care,

14(5), 377–391.

Hollingsworth D., 1995. TheWorkflow Reference Model.

Technical report wfmctc- 1003, 1995, Workflow

Management Coalition.

Jebalia S. C., 2008. Modélisation et analyse de

l’organisation du fonctionnement des structures

d’hospitalisation à domicile. Thèse de doctorat,

l’Ecole Centrale des Arts et Manufactures Spécialité :

Génie Industriel.

Ben Bachouch R., Guinet A., Hajri-Gabouj S., 2009.

Certification des HAD VS certification ISO 9000.

Congrès International de Génie Industriel (CIGI09),

Bagnère de Bigorre, France, 10-12 Juin, Actes

CDROM.

Lunn K., Sixsmith A., Lindsay A., Vaarama M., 2003.

Traceability in requirements through process

modelling, applied to social care applications.

Information and Software Technology, 2003, p. 1045-

1052.

Hamek S., Anceaux F., SOUF N., Flahou S., 2005. La

prise en charge des soins lors de l’hospitalisation à

domicile : une analyse de la coopération asynchrone.

Journées Francophones d’Informatique Médicale.

FNEHAD:Fédération nationale des établissements

d'hospitalisation à domicile "Livre blan des systèmes

d'information en hospitalisation à domicile "Juin 2009.

ISO (2000). "International Organization for

Standardization:ISO 9000:2000: Quality Management

Systems" – Fundamentals and Vocabulary.

Beuth,Berlin.

Dadam P., Reichert M., Kuhn K., 2000. Clinical

Workflows - The Killer Application for Process-

oriented Information Systems? Proc. of Int'I

Conference on Business Information Systems, Pozan,

Poland, April 2000, Springer-Verlag, 2000, pp.36-59.

Dazzil L., Fassino C., Saracco R., Quaglini S., Stefanelli

M., 1997. A Patient Workflow Management System

Built on Guidelines. Proc. of AMIA, 146-150,

Nashville, TN.

Quaglinia S., Stefanellia M., Lanzolaa G., Caporussob V.,

Panzarasab S., 2001. Flexible guideline-based patient

careflow systems. Artificial Intelligence in Medicine

22.

Arbaoui S., Lamine E., Di Mascolo M., 2008. Les

systèmes d’information pour la prise en charge de

patients à domicile: Besoins, Usages et Verrous. 4e

conférence francophone en Gestion et Ingénierie des

systèmes hospitaliers (GISEH), ISBN 978-2-8399-

0316-5, CD-ROM,Lausanne.

Cunin P., 2000. The PIE Project: An Introduction. In

Proceedings of the 7th European Workshop in

Software Process Technology – EWSPT’2000,

Kaprun, Austria.

Liu C., Li Q., Zhao X., 2008. Challenges and opportunities

in collaborative business process management:

Overview of recent advances and introduction to the

special issue. Springer Science and Business Media,

LLC.

Thomas R. Gruber., 1993. Towards Principles for the

Design of Ontologies Used for Knowledge Sharing. In

N. Guarino and R. Poli, editors, Formal Ontology in

Conceptual Analysis and Knowledge Representation,

Deventer, The Netherlands, 1993. Kluwer Academic

Publishers.

Valls A., Gibert K., Sánchez D., Batet M., 2010. Using

ontologies for structuring organizational knowledge in

Home Care assistance. In international journal of

medical informatics 7 9 (2010) 370.

Ghidini, C. D., Francescomarino, M. Rospocher, P.

Tonella,and L. Seraﬁni, 2011. Semantics based aspect

oriented manage-ment of exceptional ﬂows in business

processes. IEEE Transactions on Systems, Man, and

Cybernetics, Part C:Applications and Reviews.

Zarour, K., and Zarour, N., 2007. Vers un système

d'information coopératif pour la prise en charge des

soins à domicile: une architecture basé agent. Memoire

de magister, Université Mentouri de Constantine,

laboratoire LIRE.

Uschold, M., and King, M., 1995. Towards a

Methodology for Building Ontologies. In D. Skuce

(Ed.), IJCAI’95 Workshop on Basic Ontological Issues

in Knowledge Sharing, (pp. 6.1-6.10). Montreal,

Canada.

Staab, S., Schnurr, H. P., Studer, R., and Sure, Y., 2001.

Knowledge Processes and Ontologies. IEEE

Intelligent Systems, 16 (1), 26–34.

Song X., Hwong B., Matos G., Rudorfer A., Nelson C.,

Han M., Girenkov A., 2006. Understanding

Requirements for Computer-Aided Healthcare

Workflows: Experiences and Challenges. ICSE’06,

May 20-28, 2006, Shanghai, China. ACM 1-59593-

085-X/06/0005.

Bouzguenda L., Bouaziz R., Andonoff E., 2008. Using

ontologies for coordination in loose Inter-

Organizational Workflow. RCIS, 2008, p. 123-132.

Vieira, T. A. S. C., M. A. Casanova, and L. G. Ferrao,

2004. An Ontology-Driven Architecture for Flexible

WorkflowExecution. In WebMedia & LA-Web Joint

Conference. 2004.

HEALTHINF 2012 - International Conference on Health Informatics

236