User Proﬁling of People with Disabilities

A Proposal to Pervasively Assess Quality of Life

Eloisa Vargiu

, Luigi Ceccaroni

, Laia Subirats

, Suzanne Martin

and Felip Miralles

Barcelona Digital Technology Center, C/ Roc Boronat 117, Barcelona, Spain

Health & Rehabilitation Sciences Research Centre, University of Ulster, Jordanstown, Northen Ireland, U.K.

Keywords:

User Proﬁle, Quality of Life Assessment, Sensor-based Telemonitoring, Home Support, BNCI.

Abstract:

This paper presents and discusses an ongoing work aimed at deﬁning the proﬁle of people with disabilities,

i.e., automatically assessing their quality of life, through a sensor-based telemonitoring system. To illustrate

how the approach works, a case study is presented and discussed.

1 INTRODUCTION

Proﬁling users comes from the need to acquire infor-

mation about habits, preferences, tastes as well as de-

mographic data of the users of a service or a system

(Godoy and Amandi, 2005) (Armano et al., 2010).

In healthcare domain, proﬁling users helps health-

care providers, medical doctors, and carers to improve

their knowledge about the patients and personalize

treatments and tasks, accordingly. In our view, a rele-

vant part of the proﬁle of people with health diseases

is their Quality of Life (QoL); i.e., the subjective ex-

periences or preferences expressed by an individual

in relation to speciﬁed aspects of her/his QoL, with

a particular reference to the health status (Sutherland

and Till, 1993). Therefore, we propose to use the as-

sessment of QoL of patients as a proxy of the deﬁni-

tion of their proﬁle.

Currently, to assess QoL, several questionnaires

have been proposed and adopted; users are asked to

answer to a predeﬁned set of questions about their

mental and psychological status and feeling. Al-

though they are largely adopted, answering them

could become tedious and annoying for users. An-

swering them could also be impossible in cases of se-

vere impairment of the user. Hence, other solutions

need to be investigated. Telemonitoring is a way to

assess QoL at distance (Clark et al., 2007). Telemoni-

toring can be useful to collect information on patients’

health-status parameters that provide a measure of

their QoL and their level of functional diversity, tak-

ing into account not only functional and cognitivefac-

tors, but also psychological, social, and participation

ones.

In this paper, we present a proposal to pervasively

assess QoL automatically gathering standardized in-

formation by a sensor-based telemonitoring system.

This proposal is part of the research carried out within

BackHome

, an EU project about physical and so-

cial autonomy of people with functional diversity,

in which Brain-Neural Computer Interfaces (BNCIs)

and other assistive technologies are used. BackHome

supports the transitions from hospital care to post re-

habilitation at home to discharge (Daly et al., 2012).

The rest of the paper is organized as follows. Sec-

tion 2 gives a viewof the background. In section 3, we

describe a scenario aimed at highlighting how the ap-

proach illustrated in Section 4 can be put in practice.

Section 5 presents the proposed case study. Section 6

ends the paper with conclusions and future work.

2 QUALITY OF LIFE

ASSESSMENT

2.1 Deﬁnitions

QoL (sometimes refereed to as Health-Related QoL

or HRQoL) is deﬁned by the subjective experiences

or preferences expressed by an individual, or mem-

bers of a particular group of persons, in relation to

speciﬁed aspects of health status that are meaning-

ful, in deﬁnable ways, for that individual or group

(Sutherland and Till, 1993). According to (Gotay and

Moore, 1992), QoL is a state of well-being deﬁned

http://www.backhome-fp7.eu/backhome/index.php

352

Vargiu E., Ceccaroni L., Subirats L., Martin S. and Miralles F..

User Proﬁling of People with Disabilities - A Proposal to Pervasively Assess Quality of Life.

DOI: 10.5220/0004330003520357

In Proceedings of the 5th International Conference on Agents and Artiﬁcial Intelligence (ICAART-2013), pages 352-357

ISBN: 978-989-8565-39-6

 2013 SCITEPRESS (Science and Technology Publications, Lda.)

by two components: (i) the ability to perform every-

day activities, which reﬂects physical, psychological

and social well-being, and patient satisfaction with

levels of functioning, and (ii) the control of disease

and treatment symptoms. Also, as Lerer (2000) sug-

gests, e-health consumers are now empowered by an

increased ability to obtain health information via the

Internet, with the main objective to maintain the high-

est possible level of QoL.

The World Health Organization (WHO) deﬁnes

QoL as the individuals’ perception on their position

in life within the cultural context and the value sys-

tem in which the individuals live and with respect to

their goals, expectations, norms and worries (WHO,

2007). It is a multidimensional and complex concept

that includes personal aspects, like health, autonomy,

independence, satisfaction with life and environmen-

tal aspects such as support networks and social ser-

vices, among others.

QoL could also be considered as a dynamic and

changing concept that includes continuous interac-

tions between the person and the environment. Ac-

cordingly, QoL in ill people is related to the interac-

tion among the disease, the patients’ character, the

change in their life, the received social support, as

well as the period of life in which the disease appears.

Healthcare organizations use several tools to ac-

quire QoL-related information. These tools make use

of speciﬁc terms, which are sometimes ambiguous:

descriptor, grade, index, indicator, parameter, ques-

tionnaire, scale, score, and test. The terminology used

in this paper is deﬁned as follows and is part of an

ontology, which we deﬁned (and encoded in OWL 2

(Hitzler et al., 2009)) based on standard nomencla-

tures and ontologies (Ceccaroni and Subirats, 2012):

• Indicator: a (subjective or objective) parameter,

category, or descriptor used to measure or com-

pare activities and participation, body functions,

body structures, environment factors, processes,

and results (e.g., dressing).

• Index: a combination of indicators, questionnaires

and possibly other indexes. The function repre-

senting this combination gives as summarizing re-

sult a score (e.g., Barthel index).

• Item: a single question or concept (e.g., Mobility).

• Questionnaire (or instrument or test): a set of

questions (or items) answered using a scale (e.g.,

EQ-5D).

• Scale: a mapping between some ordered (qualita-

tive or quantitative) values (or grades) and their

description. These values are used to answer

questionnaires (e.g., I have no problems in walk-

ing about, I have some problems in walking about,

I am conﬁned to bed).

2.2 Questionnaires for Assessment of

Quality of Life

Several questionnaires have been proposed and

adopted to assess QoL. Let us summarize here the

most widely adopted:

• The WHOQOL-BREF questionnaire (Murphy

et al., 2000) comprises 26 items, which measure

the following broad domains: physical health,

psychological health, social relationships, and en-

vironment.

• The EQ-5D-5L questionnaire (The Euroqol

Group, 1990) was developed by the EuroQol

Group in order to provide a generic measure of

health status. Applicable to a wide range of health

conditions and treatments, it providesa simple de-

scriptive proﬁle and a single value for health sta-

tus that can be used in the clinical and economic

evaluation of healthcare as well as in population

health surveys.

• The RAND-36 questionnaire (Hays et al., 2001)

is comprised of 36 items that assess eight health

concepts: physical functioning, role limitations

caused by physical health problems, role limita-

tions caused by emotional problems, social func-

tioning, emotional well-being, energy/fatigue,

pain, and general health perceptions.

• The Short Form (36) Health Survey (SF-36v2)

(Ware et al., 2001) is a questionnaire about patient

health status and is commonly used in health eco-

nomics in the quality-adjusted life year calcula-

tion to determine the cost-effectivenessof a health

treatment. The SF-36 and RAND-36 include the

same set of items, however the scoring of general

health and pain is different (RAND et al., 1992).

• The Barthel questionnaire (O’Sullivan and

Schmitz, 2007) is used to measure performance

in Activities of Daily Living (ADLs). It uses ten

variables describing ADLs and mobility. The

higher the score derived from this questionnaire,

the greater the likelihood of being able to live

at home with independence following discharge

from hospital.

2.3 Existing Standardization Efforts

Several standard terminologies and classiﬁcations ex-

ist, which can be used for an interoperable rep-

resentation of QoL. Some examples are: the Sys-

tematized Nomenclature of Medicine Clinical Terms

UserProfilingofPeoplewithDisabilities-AProposaltoPervasivelyAssessQualityofLife

353

(SNOMED CT); the Uniﬁed Medical Language Sys-

tem (UMLS); the International Classiﬁcation of Dis-

eases version 10 (ICD-10); and the International Clas-

siﬁcation of Functioning, disability and health (ICF)

deﬁned by the WHO. In addition to terminologies and

classiﬁcations, information models such as the virtual

Medical Record (vMR) contribute to solve interoper-

ability problems in the electronic exchange of QoL

information.

Several questionnaires are used to evaluate func-

tioning, disability and health. The ICF classiﬁes these

concepts, speciﬁes their range of values, and can be

used to solve interoperability problems among health

institutions that employ different measuring question-

naires. To this aim, questionnaire items can be en-

coded to ICF concepts following the standardization

methodology proposed by Cieza et al. (2005). Difﬁ-

culties in mapping clinical questionnaires to standard

terminologies and ontologies in the rehabilitation do-

main (e.g., data from questionnaires having a ﬁner

granularity than ICF categories) have been addressed

in (Ceccaroni and Subirats, 2012) and (Subirats et al.,

2012). ICF core sets are subsets of the ICF that have

been created according to speciﬁc pathologies or re-

habilitation processes. Core sets are useful because,

in daily practice, clinicians and other professionals

can use only a fraction of the about 1400 categories

found in the ICF.

3 A REFERENCE SCENARIO

The next real case may illustrate the urgent need for

automatic, interoperable assessment of QoL for peo-

ple with disabilities caused by neurological disor-

ders

Chara is a woman in her thirties who used to be

a painter. She was so gifted that she could earn a liv-

ing from painting. About eight years ago she started

to have difﬁculties holding her paint brush. She was

diagnosed with Amyotrophic Lateral Sclerosis (ALS).

She was devastated because it meant that in the near

future she would no longer be able to paint. She be-

came tetraplegic and artiﬁcially ventilated, and she

was so depressed that she refused treatment when di-

agnosed with pneumonia. Luckily, she survived and

just recently had her ﬁrst session with BNCI-based

Brain Painting (M

unssinger et al., 2010). We let her

speak: “Here is my feedback to my ﬁrst brain painting

image. I am deeply moved to tears. I have not been

able to paint for more than ﬁve years. Today I again

had butterﬂies in my stomach, a feeling that I have

Names have been changed for privacy reasons.

missed so much. I was so sad; I was plagued by fears

of loss. For me the picture I have created is so very

typical me, no one else paints in my style, and despite

ﬁve years of absence, I’m simply an artist again. I’m

back to life!”

To achieve an automatic and interoperable QoL

assessment for people like Chara, several actions need

to be taken, as part of our proposal.

Before her discharge from the hospital, therapists

ask her to answer some questions in order to have an

initial assessment of her QoL according to selected

standards

. At home, Chara, with the help of a care-

giver or of her relatives (but no BNCI experts), is

able to put/remove a BNCI system as well as other

wearable sensors. In so doing, she can: control a

sensor-based smart home, participate in social net-

works, carry out cognitive rehabilitation, and inter-

act with a telemonitoring system. The telemonitor-

ing system acquires all relevant data about Chara and

transmits them to a remote therapist for future anal-

ysis. In particular, the system monitors the follow-

ing daily activities: interaction with home automation

devices, communication with relatives and friends,

movement around home, and the rehabilitation tasks

assigned by the therapist. The telemonitoring system

continuously analyses the data gathered by the sen-

sors in order to send alarms in case of detection of ab-

normal events, if any, and to keep informed the ther-

apist about the performed activities. Furthermore, the

system analyses all the data to assess the QoL. In case

of QoL worsening, the system sends an alarm to the

therapist, who might see it necessary to take action or

to change the therapy.

4 PERVASIVE ASSESSMENT OF

QUALITY OF LIFE

As already mentioned, currently, approaches to assess

QoL rely on questionnaires, which the user is peri-

odically required to answer. The more the user ﬁlls

the questionnaire, the more the QoL trend is updated.

Moreover, the more the questions, the more the QoL

trend is accurate. Unfortunately, this process could

become tedious and annoying for the users, especially

if they are asked to do that frequently; with the poten-

tial consequence of users who stop answering ques-

tionnaires or give less accurate answers. To take into

account these issues we propose, after a basic proﬁle

has been deﬁned, to adopt a sensor-based telemoni-

toring system to automatically ﬁll it.

This is a one-time questionnaire.

ICAART2013-InternationalConferenceonAgentsandArtificialIntelligence

354

4.1 The Approach

The proposed approach falls in the domain of Am-

bient Assisted Living, which fosters the provision of

equipment and services for the independent or more

autonomous living of people with disabilities, via

the seamless integration of info-communication tech-

nologies within homes and residences, thus increas-

ing their physical and social autonomy and reducing

the need for hospital readmission. In particular, the

proposed approach is aimed at acquiring personalized

information through data coming from: (i) a BNCI

system that allows monitoring ElectroEncephalo-

Gram (EEG), ElectroOculoGram (EOG), and Elec-

troMyoGram (EMG); (ii) wearable, physiological,

and biometric sensors, such as ElectroCardioGram

(ECG), heart-rate sensor, respiration-rate sensor, Gal-

vanic Skin Response (GSR) sensor, EMG switches,

Wii remote, and inertial sensors (e.g., accelerometer,

gyrocompass, and magnetometer); (iii) environmental

sensors (e.g., temperature and humidity sensors); (iv)

smart home devices (e.g., wheelchairs, lights, TVs,

doors, windows and shutters); (v) devices that allow

interaction activities (e.g., a desktop PC); and (vi) de-

vices to perform rehabilitation tasks (e.g., a robot).

In this way, the system is able to monitor the evo-

lution of the user’s daily life activity at home, once

discharged from the hospital Vargiu et al. (2012).

Speciﬁcally, the wearable sensors allow to monitor fa-

tigue, spasticity, stress, and further user’s conditions.

Environmental sensors are used to monitor tempera-

ture and humidity, as well as the movements (motion

sensors) and the physical position of the user (loca-

tion sensors). Smart home devices enable physical

autonomy of the user and help her/him carry out daily

life activities. An Internet-connected device allows

the user to communicate with remote therapists, ca-

reers, relatives, and friends through Skype, email, or

social networks (i.e., Facebook and Twitter). Other

devices that allow some kind of interaction and stim-

ulation activities are also taken into account, such as

devices to play games, listen to music, paint, or per-

form other activities.

4.2 Kinds of Data

If the user explicitly agrees, the logs of communica-

tion, social, and all other activities are used to study

the behaviour and the mood of the user. In this way,

the system is able to pervasively assess the QoL of

the user automatically ﬁlling a selected standard ques-

tionnaire. Two kinds of data are considered: moni-

torable and inferable. On the one hand, all data that

can be gathered from the wearable, home automa-

tion, and environmental sensors, as well as the BNCI

system (i.e., without relying on direct input from the

user) belong to the former category. This kind of data

allows, for instance, to answer the question “How

well are you able to get around?”

. On the other

hand, all data inferred by analysing data retrieved by

the system (e.g., by considering activities performed

by the user while interacting with a social network)

belong to the latter category. This kind of data allows,

for instance, to answer the question “How satisﬁed

are you with yourself?”

Let us note that this does not imply that moni-

torable and inferable data are necessarily monitored

or inferred. In fact, in BackHome we decided to not

monitor nor infer some data (such as, those related

to self-care and sleeping activities), due to privacy is-

sues. Moreover, users can decide to switch off the

monitoring of any descriptor.

5 A CASE STUDY

Among the several state-of-the-art questionnaires, ac-

cording to the constraints of BackHome, we made

a preliminary study by considering the EQ-5D-5L

questionnaire (The Euroqol Group, 1990)

. The EQ-

5D-5L is a brief, self-administered, two-page ques-

tionnaire. The ﬁrst page contains ﬁve items describ-

ing health status across ﬁve dimensions: mobility,

self-care, usual activity, pain/discomfort, and depres-

sion/anxiety. The second page has a visual analogue

rating scale on which the respondent marks an as-

sessment of her/his overall health. Each dimension

is divided into ﬁve levels which, when taken together,

deﬁne a total of 3125 (5

) unique health states. The

responses to the ﬁve items in the EQ-5D-5L can be

scored using a utility-weighted algorithm (Williams

et al., 1995), which has been recommended for use in

economic evaluation. The EQ-5D-5L, therefore, pro-

vides two single-index measures of health, the rating

scale, and the EQ-5D index, ranging from 0 to 100

(Brooks, 1996).

According toGeyh et al. (2007), Table 1 shows the

translation of the selected questionnaire into the ICF

categories. The visual analogue rating scale, which

refers to how good or bad the user feels her/his health

is “today”, can be also translated on “emotional func-

tions”.

The ICF classiﬁcation allows us to “zoom in” in

each EQ-5D-5L dimension and better study its corre-

spondence with the data introduced in Section 4.2:

Item number 15 of the WHOQOL questionnaire

Item number 19 of the WHOQOL questionnaire

An ad-hoc questionnaire is currently under deﬁnition.

UserProfilingofPeoplewithDisabilities-AProposaltoPervasivelyAssessQualityofLife

355

Table 1: The translation of the EQ-5D dimensions into the ICF categories.

EQ-5D Dimension ICF Category

Mobility

d4 - mobility

d450 - walking

d498 - mobility, other speciﬁed

Self-care

d5 - self-care

d510 - washing oneself

d540 - toileting

d540 - dressing

Usual Activities

d6409 - doing housework, unspeciﬁed

d7609 - family relationships, unspeciﬁed

d839 - education, other speciﬁed and unspeciﬁed

d8509 - remunerative employment, unspeciﬁed

d9209 - recreation and leisure, unspeciﬁed

Pain/Discomfort

b152 - emotional functions

b280 - sensation of pain

b289 - sensation of pain, other speciﬁed and unspeciﬁed

Anxiety/Depression b152 - emotional functions

• Monitorable Data

– Mobility. Through the adoption of environ-

mental location sensors, the system is able to

know the position of the user, time after time. It

is worth pointing out that, in BackHome, users

are typically on a wheelchair, thus the walk-

ing activity is not of interest here. To detect

the position of the wheelchair and its move-

ments RFID tags will be embedded into the

wheelchair together with following sensors.

– Usual Activities. Being human-computer-

interaction made through a BNCI system, it is

possible to monitor all the activities performed

by the user on the PC and while interacting

with smart home control and communication

devices. In other words, the system is able,

through the BNCI system, to know which ac-

tion is performed.

• Inferable Data

– Usual Activities. The user can interact with

her/his family and friends through the support

of a communication system (e.g., Skype) or

social network (e.g., Facebook and Twitter).

Thus, suitable text mining algorithms can be

adopted to infer the family and friends relation-

ship status.

– Pain/Discomfort. Text mining algorithms, ap-

plied on social networking and communication

activities, can also be adopted to assess the de-

gree of pain or discomfort.

– Anxiety/Depression. Changes observed in

habits of daily life activities can be studied

to assess anxiety or depression. The degree

of anxiety/depression can be also inferred by

analysing data on fatigue, spasticity, stress,

and further user’s conditions retrieved by the

BNCI system and the other wearable sensors.

Moreover, analogously to pain and discomfort,

anxiety and depression can be inferred by the

system by adopting suitable text mining algo-

rithms on the performed social activities.

In principle, also self-care activities can be moni-

tored by relying on suitable sensors. Nevertheless, in

BackHome, for privacy issues, the end-user decided

not to monitor such activities. In so doing, to have

information about the self-care activities and the cor-

responding status of the user, direct questions must be

performed.

6 CONCLUSIONS

Being interested in proﬁling people with disabilities

by assessing their QoL, in this paper we proposed

an approach to automatically assess QoL by relying

on a sensor-based telemonitoring system. A case

study, which relies on the EQ-5D-5L questionnaire,

has been presented. The case study allows us to clar-

ify how a given questionnaire can be automatically

ﬁlled by relying on a sensor-based telemonitoringsys-

tem.

This is an ongoing work; we are currently setting

up the sensor-based telemonitoring system and we are

also selecting, according to the inclusion and exclu-

sion criteria deﬁned in BackHome, the end-users who

will test the overall system and who will be asked

ICAART2013-InternationalConferenceonAgentsandArtificialIntelligence

356

to test the proposed pervasive QoL assessment ap-

proach. Finally, we are planning to assess the validity

of the proposed approach with respect to the tradi-

tional ﬁlling of questionnaire.

ACKNOWLEDGEMENTS

The research leading to these results has received

funding from the European Community’s, Seventh

Framework Programme FP7/2007-2013, BackHome

project grant agreement n. 288566.

REFERENCES

Armano, G., de Gemmis, M., Semeraro, G., and Vargiu, E.

(2010). Intelligent Information Access, volume SCI 301.

Springer-Verlag, Studies in Computational Intelligence

series, Heidelberg, Germany.

Brooks, R. (1996). EuroQol: the current state of play.

Health Policy, 37(1):53–72.

Ceccaroni, L. and Subirats, L. (2012). Interoperable knowl-

edge representation in clinical decision support systems

for rehabilitation. International Journal of Applied and

Computational Mathematics, 11(2):303–316.

Cieza, A., Geyh, S., Chatterji, S., Kostanjsek, N.,

Ust¨un, B.,

and Stucki, G. (2005). Icf linking rules: an update based

on lessons learned. Rehabil. Med., 37:212–218.

Clark, R. A., Inglis, S. C., McAlister, F. A., Cleland, J. G. F.,

and Stewart, S. (2007). Telemonitoring or structured

telephone support programmes for patients with chronic

heart failure: systematic review and meta-analysis. BMJ,

334(7600):942.

Daly, J., Armstrong, E., Miralles, F., Vargiu, E., M¨uller-

Putz, G., Hintermller, C., Guger, C., Kuebler, A., and

Martin, S. (2012). BackHome: Brain-neural-computer

interfaces on track to home. In RAatE 2012 - Recent

Advances in Assistive Technology & Engineering.

Geyh, S., Cieza, A., Kollerits, B., Grimby, G., and Stucki,

G. (2007). Content comparison of health-related quality

of life measures used in stroke based on the international

classiﬁcation of functioning, disability and health (icf): a

systematic review. Quality of Life Research, 16(5):833–

851.

Godoy, D. and Amandi, A. (2005). User proﬁling in per-

sonal information agents: a survey. Knowl. Eng. Rev.,

20(4):329–361.

Gotay, C. and Moore, T. (1992). Assessing quality of life in

head and neck cancer. Qual Life Res., 1:5–17.

Hays, R., Morales, L., RAND Health, P., and RAND, C.

(2001). The RAND-36 Measure of Health-related Qual-

ity of Life. RAND Health reprint. RAND.

Hitzler, P., Kr¨otzsch, M., Parsia, B., Patel-Schneider, P. F.,

and Rudolph, S., editors (2009). OWL 2 Web Ontology

Language: Primer. W3C Recommendation.

Lerer, L. (2000). The healthcare 2020 platform: The e-

health consumer. PhD thesis, INSEAD.

M¨unssinger, J., Halder, S., Kleih, S., Furdea, A., Raco, V.,

H¨osle, A., and K¨ubler, A. (2010). Brain painting: First

evaluation of a new braincomputer interface application

with als-patients and healthy volunteers. Front Neurosci,

4:182. doi: 10.3389/fnins.2010.00182.

Murphy, B., Herrman, H., Hawthorne, G., Pinzone, T., and

Evert, H. (2000). Australian WHOQoL instruments:

User’s manual and interpretation guide. Australian

WHOQoL Field Study Centre. Melbourne, Australia.

O’Sullivan, S. and Schmitz, T. (2007). Physical Rehabili-

tation. G - Reference, Information and Interdisciplinary

Subjects Series. F.A. Davis.

RAND, C., Health Sciences, P., Hays, R., Sherbourne, C.,

and Mazel, R. (1992). RAND 36-item Health Survey

1.0: RAND Health Sciences Program. RAND reprints.

RAND.

Subirats, L., Ceccaroni, L., and Miralles, F. (2012). Knowl-

edge representation for prognosis of health status in re-

habilitation. Future Internet, 4(3):762–775.

Sutherland, H. and Till, J. (1993). Quality of life assess-

ments and levels of decision making: differentiating ob-

jectives. Qual Life Res, 2(4):297–303.

The Euroqol Group (1990). Euroqol a facility for the mea-

surement of health-related quality of life. Health Policy,

16:199–208.

Vargiu, E., Miralles, F., Martin, S., and Markey, D. (2012).

BackHome: Assisting and telemonitoring people with

disabilities. In RAatE 2012 - Recent Advances in As-

sistive Technology & Engineering.

Ware, J., Kosinski, M., and Dewey, J. (2001). How to Score

Version 2 of the SF-36 Health Survey: Standars & Acute

Forms. QualityMetric.

WHO (2007). Global Age-Friendly Cities: A Guide. Non-

serial Publication. World Health Organization.

Williams, A., of York. Centre for Health Economics, U.,

Consortium, H. E., and of York. NHS Centre for Re-

views & Dissemination, U. (1995). The Measurement

and Valuation of Health: A Chronicle. Centre for Health

Economics discussion paper. University of York, Centre

for Health Economics.

UserProfilingofPeoplewithDisabilities-AProposaltoPervasivelyAssessQualityofLife

357