Pervasive Information Systems to Intensive Care Medicine

Technology Acceptance Model

Jorge Aguiar

, Filipe Portela

, Manuel Filipe Santos

, José Machado

, António Abelha

Álvaro Silva

, Fernando Rua

and Filipe Pinto

Algoritmi Centre, University of Minho, Azurém, Guimarães, Portugal

CCTC, University of Minho, Braga, Guimarães, Portugal

Serviço de Cuidados Intensivos, Centro Hospitalar do Porto, Porto, Portugal

Escola Superior de Tecnologia e Gestão, Instituto Politécnico de Leiria, Leiria, Portugal

Keywords: Information Systems, Technology Acceptance Model, Delphi, Intensive Care.

Abstract: The usability of information systems in critical environments like Intensive Care Units (ICU) is far than the

expected and desirable. Typically, ICUs have a set of not integrated information silos and a high number of

data recorded in paper. Whenever ICU professionals need to make a decision they have to deal with a high

number of data sources containing useful information. Unfortunately, they can't use those sources due to the

difficulty of evaluating them in a correct time. Pervasive Intelligent Decision Support Systems (PIDSS),

operating automatically and in real-time, can be used to improve the decision making if they are suited to

the requirements of the ICU. In this work a PIDSS have been assessed in terms of quality and user

acceptance making use of Technology Acceptance Model (TAM). TAM proved to be very useful when

combined with Delphi method features to involve the professionals and to make the system usable.

1 INTRODUCTION

Nowadays, intensive care professionals face

important obstacles to take decisions in a short time.

Intensive Care Units (ICU) are recognized as a place

where there are, a high number of electronic devices

to collect the patient data, a high number of

information silos and a high number of information

in paper. This situation contributes to increase the

number of data available in the moment of the

decision. However, to a human it is very difficult

access to all the information in a correct time

without technology help. To give support to ICU

professionals in the Decision Making process, a

research project called INTCare was developed. The

first goal of the project was to develop an Intelligent

Decision Support System (IDSS) to predict organ

failure and patient outcome for the next hour. Later,

a Pervasive IDSS (PIDSS) has been deployed. This

implied a set of modifications in the ICU

information systems (IS). In this context, was

developed an automatic and real-time data

acquisition system and a platform to record /

validate / consult the patient data in real-time. To

assure the success of information systems and

associated technologies it is very important to assess

the system quality and the acceptance level by the

users. In order to assess the technology acceptance

by the users (nurses) was used the Technology

Acceptance Model III combined with some features

of Delphi methodology. This paper provides an

overview of the system, presents the methodologies

used and the results achieved. The present work

allows a better comprehension of the importance that

a PIDSS has to the ICU needs and the respective

technology acceptance by the professionals.

This paper is divided in six sections. The first

one makes an introduction to the subject. The second

section presents some background and related work.

Then the third section presents the questionnaire

elaborated and the relationship with TAM. Fourth

section presents the results at level of TAM III.

Finally, in section five the results are discussed and

in section six some conclusions are done.

177

Aguiar J., Portela F., Santos M., Machado J., Abelha A., Silva Á., Rua F. and Pinto F..

Pervasive Information Systems to Intensive Care Medicine - Technology Acceptance Model.

DOI: 10.5220/0004441001770184

In Proceedings of the 15th International Conference on Enterprise Information Systems (ICEIS-2013), pages 177-184

ISBN: 978-989-8565-59-4

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

2 BACKGROUND

2.1 Pervasive Information System

Over the last three years, some important changes

were made in the ICU Information System (IS):

gathering and processing the data in real-time and

introduction of intelligent agents in order to perform

some tasks automatically, replacing some manual

operations. During the development of the project

the environment also was changed. This system

meets some of pervasive health care (Varshney,

2009) and pervasive computing (Orwat et al., 2008);

(Saha and Mukherjee, 2003) features, namely: health

care for anyone anywhere and anytime, remove

restraints of time and location, increase both the

coverage and the quality of healthcare; scalability,

heterogeneity, integration, invisibility and context

awareness. Now, the IS interoperates with the data

acquisition system to automatically and in real-time

provide a set of data anywhere and anytime.

Currently, the INTCare system performs all tasks

automatically, online and in a real-time.

2.1.1 Architecture

The IS architecture of ICU is divided into two

subsystems: one to collect the data and another to

process and obtain the knowledge. To obtain the

data there are two ways of acquisition: manually and

automatically. To produce the knowledge, a set of

intelligent agents are used in order to automatically

execute some tasks according to the targets (e.g.,

predict organ failure, score ICU measures, and

calculate Critical Events). Now, the data is acquired

online, in real-time and in an electronic format using

automatic or manual procedures. The data is

provided from several data sources (eg. bedside

monitors (vital signs), laboratory results, eletronic

health record (EHR), pharmacy (drugs

prescription)). Then, the data acquired are stored

into the database and made available online through

the Electronic Nursing Record (ENR).

2.1.2 Electronic Nursing Record

Electronic Nursing Record (ENR) is a platform that

was developed with the objective to collect all the

clinical data and make it available to the doctors and

nurses in an hourly-based mode. Now, using ENR,

the ICU professionals can have more information

about the patient, essential to make their decisions.

ENR is a touch and web-based platform. ENR is the

main technology of the ICU and was assessed in

terms of user acceptance. A set of questionnaires

were made having in consideration each component

of the platform. All system features were evaluated.

2.1.3 Decision Making Process

The Decision Making Process (DMP) in ICU is a

crucial process, because the professionals are

dealing with patients in critical condition. The

decision needs to be performed quickly and

assertively. Due to the high number of data sources

present in the ICU it is difficult to have a correct

decision in the right moment. In order to avoid this

problem, INTCare changed the way the data is

presented to the decision makers. Now, taking

advantage from the environment changes and using

an inference engine, new knowledge is provided in

the moment of the decision is made. The utility and

importance of this new knowledge was assessed by

TAM 3. INTCare delivers knowledge essential to

the DMP anywhere and anytime. INTCare can

provide:

 Patient Clinical data;

 Critical Events tracking;

 ICU Medical Scores;

 Probability associated to organ failure or death.

2.1.4 Intcare

INTCare (Gago et al., 2006); (Santos et al., 2011) is

an Intelligent Decision Support System (IDSS) to

predict organ failure and patient outcome in real-

time using online-learning. The work deployed

allows for obtaining new types of data electronically

and in real-time (Portela et al., 2010). New

knowledge fundamental to the decision process is

now available automatically and in real-time

(Portela et al., 2011). INTCare uses ENR to acquire

data and present some new knowledge generated in

a pervasive way, i.e., anywhere and anytime.

Intelligent agents are used for processing and

transforming the data automatically, without human

intervention in order to prepare the input variables

for the models.

2.2 Technology Acceptance Model

The evaluation of a certain technology is crucial to

understand its suitability in a specific environment

and also to measure the users’ satisfactoriness level.

One of the most used models in this area is the

Technology Acceptance Method (TAM). The main

purpose of TAM is to present an approach to study

the effects of external variables towards people’s

internal beliefs, attitudes, and intentions

ICEIS2013-15thInternationalConferenceonEnterpriseInformationSystems

178

(Chooprayoon and Fung, 2010). This model is also

important because gives an understanding about the

acceptance of modifications made in the decision

support, and how they can be useful in the course of

ICU professionals’ daily work. More recently,

Venkatesh and Bala set the TAM 2 (Venkatesh and

Bala, 2008) to a model using determinants of the

perceived ease of use(Venkatesh and Davis, 2000)

and developed an integrated model nominated TAM

3. TAM 3 is composed by four constructs: Perceived

Ease of USE (PEOU), Perceived Usefulness (PU),

Behaviour Intention (BI) and Use Behaviour (UB)

that are derived from other type of analysis as can be

seen in Figure 2.

Figure 1: Technology Acceptance Model 3 (TAM 3)

(Venkatesh and Bala, 2008).

In order to obtain plausible results in the analysis

of the questionnaires, a program called statistical

data analysis by PAleontological STatistics (PAST)

was used (Hammer et al., 2001). To evaluate the

correlation results was used Kendall’s tau technique.

This technique is a non-parametric correlation

coefficient that can be used to assess and test

correlations between non-intervals scaled ordinal

variables. Kendall’s tau (Bolboaca and Jantschi,

2006) is used as a statistical test to determine if two

variables can be considered as statistically

dependent. The correlation coefficient should deliver

a range of [-1, 1]. If the agreement between the two

evaluations is perfect, the coefficient has a value (1).

If the divergence between the two evaluations is

perfect (inverse of the other), the coefficient has a

value (-1), but if the two evaluations are

independent, the coefficient is nearly zero (Bolboaca

and Jantschi, 2006).

2.3 TAM III and Delphi

The goals of TAM can be achieved by using

methodologies based on questionnaires. As a support

tool it is important to use some aspects/characteristic

of the Delphi method. The basis of the Delphi

method involves the use of questionnaires being one

of its key features (Zackiewicz and Salles Filho,

2010), the preservation of anonymity of the

participants. The questionnaire was prepared by a

coordination team, composed by ICU and IS

professionals, and sent to participants: a group of

experts from the ICU professional team. The

questionnaire was prepared having in account the

constructs of TAM (Venkatesh and Bala, 2008);

(Venkatesh and Davis, 2000) and has as support tool

the Delphi method.

Table 1: TAM advantages and Delphi disadvantages.

Advantages TAM Disadvantages Delphi

- Important not just a technical view,

must also direct attention to the

requirements offered by technology

in order to understand user

behaviour.

- Identification of

specialist to respond to

the questionnaires

- Strand/Slope Quantitative: aims to

understand the social or human

problems from tests of existing

theory, using variables measured

with numbers and analysed with

static procedures

- In many times, the

projections that do, are

wrong or influenced

- It is useful to identify the reason for

non-acceptance of a particular

technology or system by users and

subsequently implement the

appropriate corrections

- Sometimes, they are

ambiguous and divergent

specialists in the same

area

Table 2: Delphi advantages and TAM disadvantages.

Advantages Delphi Disadvantages TAM

- Important that all related

viewpoints are represented

and pay attention to cultural

differences and cognitive

character

- Not evaluated the organizational

context in which the system is

involved, does not evaluated the

situations of centralization,

conflict, hierarchy, stability,

uncertainty of the company

- Defined as an iterative

process designed to combine

opinions of a group of

specialists to achieve a

consensus

- Difficulties in researching the

technology acceptance by the

user with all the variables

involved in its real-time

environment

- necessary to ensure

diversity in the composition

of the group of participants,

so that they cancel each

other

- Many studies use self-report

response type for the verification

of system use

Several authors point to the importance of

combining more than one method or technique on

the methodological framework of an exercise

prospective, to reduce the levels of uncertainty

inherent to type of activity, integrating more

approaches and results. When quantitative methods

(TAM) are combined with qualitative methods

PervasiveInformationSystemstoIntensiveCareMedicine-TechnologyAcceptanceModel

179

(Delphi Method), the explicit knowledge adds up to

tacit knowledge, in seeking complementary or

different views (Santos and Amaral, 2004).

Therefore, the union of the two methods may

involve an improvement in quality and greater

certainty of the results of the evaluation system to

IDSS – INTCare, since the advantages of one

method may mitigate the disadvantages of the other

method. Table 1 and 2 present these points.

3 QUESTIONNAIRES

For this study it was elaborated a questionnaire

based on four constructs of TAM 3: PEOU, PU, BI,

UB. It means that the questionnaire was aggregated

in several groups to represent all the aspects of

TAM. The questionnaire is composed by 96

questions. In this questionnaire was applied the

Likert Scale (Johns, 2010) to evaluate the results. As

a consequence, the chosen scale will follow a range

from one to five, because it gives two values for

each side and at the same time finds a neutrality

point (Johns, 2010). The considered levels were:

1) Not satisfies/in complete disagreement;

2) Satisfies a bit/in some level of disagreement;

3) Satisfies/under some level of agreement;

4) Satisfies a lot/strongly agreement;

5) Satisfies completely/full agreement.

The responses always depend on the goodwill of

each participant by answering in a balanced way to

the questions of a certain group. To avoid wrong

answers it was added three screening questions to

understand the level of the user’s consciousness (ex:

3+2). The nurses scored the questions from 1 (worst)

to 5 (best) points.

4 RESULTS

A preliminary analysis was made after collecting 14

answers (35% of total number of nurses in ICU).

After this, a deeper analysis was done to exclude

invalid or inconsistent answers given by the

participants. Only one participant out of the 14

nurses answered the questionnaire in an inconsistent

way. This situation leads us to consider as valid the

other 13 questionnaires. Table 3 presents the

technology experience of the respondents.

Table 3: Level of experience in information technology

Question Answer %

How much

time do you

spend at the

computer?

Less than 2 hours/day 0%

Between 2 to 4 hours/day 57%

More 4 hours/day 36%

Type of

User?

Full Autonomy 62%

Rarely need technical support 38%

Need regular technical support 0%

Uses

computer

preferably

for?

(multiple)

Application of production staff (email,

text processing, spread sheet)

62%

Handling/Consult administrative info 31%

Handling/Consult clinical info 77%

Handling/Consult management Info 8%

4.1 Analysis

In order to obtain plausible results in the analysis of

the questionnaires, it was necessary to use a program

called statistical data analysis by Paleontological

Statistics (PAST) (Hammer et al., 2001). A

comprehensive analysis has been performed on all

the responses, excepting the text and dispersion

questions (Ex: one + one), as well as an analysis of

the four constructs of TAM 3 by calculating:

 Mean, standard deviation (univariate);

 Correlation Coefficient (correlation);

 Bar Chart (histogram).

Finally, an analysis was performed on the results by

participant, by question, calculating the mean (bars)

and mode values (line).

4.1.1 Global Analysis

A Global Analysis of the all responses is presented

in tables 4 and 5. The nurses who participated in the

response to the questionnaires are represented by A

to M.

As can be seen in Table 4 the mean of

responses/evaluations corresponds to level 3.

Standard Deviation (STD) shows a small dispersion,

i.e. the variability in the responses is minimal. The

standard deviation is close to zero, what means that

the respondent maintained a consistency of response

(e.g. nurse (A) showed a deviation of 0, 05858516).

Table 4: Results of univariate analysis global.

The results of the correlation coefficient

ICEIS2013-15thInternationalConferenceonEnterpriseInformationSystems

180

Kendall’s tau are presented in Table 5.

Table 5: Results correlation coefficient Kendall's tau.

In Table 5 nurse G denoted a greater divergence

in the responses (values near to -1). Nurses C and H

showed a good correlation of responses, because

they have a Kendall of 0, 87. This value represents

that between them, the answers were very similar.

Trough Histogram (Figure 3), it was found that the

vast majority of respondents answered to 91

questions of the questionnaire with a rating between

3 and 4 points.

Figure 2: Global analysis histogram.

4.1.2 Analysis of the Perceived Usefulness

In order to understand which TAM 3 constructs

achieved better results, an evaluation was made for

each one of them. Table 6 presents the univariate

analysis to perceived usefulness (PU).

Table 6: Results of univariate analysis (PU).

The mean of responses/evaluations was fixed

around the three points. Standard deviation (STD)

denotes a small dispersion. For example, nurse H

maintained consistency of response, showing a

deviation of 0, 0720334. In Table 7 nurse G has a

bigger divergence response, since values were very

close to - 1.While the nurse A showed a lowest

variance with a correlation coefficient nearest to 1.

Based in the histogram of Figure 3 it can be seen

that the most respondents answered 49 questions

related to the construct Perceived Usefulness with an

evaluation positioned between 3 and 4 points.

Table 7: Results correlation coefficient Kendall's tau (PU).

Figure 3: Histogram (PU).

4.1.3 Analysis of Perceived Ease of Use

Through a brief analysis of Table 8, it was found

that the average of responses/evaluations in this

construct was approximately around four points.

Standard deviation (STD error) is not dispersed. It

can be seen that responses are consistent (e.g. Nurse

M had a deviation of 0, 05100626).

Table 8: Results of univariate analysis (PFU).

In Table 9 nurse G showed a bigger divergence

of responses, since values were very close to -1.

However, the nurse E already showed a lower

divergence with a correlation coefficient near to 1.

PervasiveInformationSystemstoIntensiveCareMedicine-TechnologyAcceptanceModel

181

Table 9: Results of the correlation coefficient Kendall's tau

(PFU).

Histogram of Figure 4 indicates that the most

respondents answered 74 questions related to the

construct Perceived Usefulness with an evaluation

located between 3 and 4 points. For example, nurse

M (blue) answered 64% of the questions with 4

points.

Figure 4: Histogram (PFU).

4.1.4 Analysis of Behavioral Intention

Table 10 indicates that only 5 nurses used all point

scales. Standard deviation (STD error) is not

dispersed. For example, nurse M showed a deviation

of 0, 07173386.

Table 10: Results of univariate analysis (BI).

In Table 11 nurse G already showed a bigger

divergence of responses, since values were very

close to -1. However, the nurse B and J showed a

lower divergence with a correlation coefficient near

to 0,86.

Histogram of Figure 5 indicates that the most

respondents answered the 41 questions related to the

construct Behavioural Intention with an evaluation

located between 3 and 4 points. For example, nurse

M (blue) answered 35 % of the questions with 4

points. At same time is possible observe that only

35% of the nurses answered questions with 1 point.

Table 11: Results correlation coefficient Kendall's tau

(BI).

Figure 5: Histogram (BI).

4.1.5 Analysis of Use Behaviour

Through a brief analysis of Table 12, it was found

that the average of responses/evaluations in this

construct was approximately around three points.

Standard deviation (STD error) is not dispersed.

Table 12: Results of univariate analysis (UB).

Table 13: Results of the correlation coefficient Kendall's

tau (UB).

ICEIS2013-15thInternationalConferenceonEnterpriseInformationSystems

182

In Table 13 nurse G and H showed a bigger

divergence of responses, since values were very

close to -1. However, the nurse A showed a lower

divergence with a correlation coefficient near to 1.

Histogram of Figure 6 indicates that the most

respondents answered 47 questions related to the

construct Use Behaviour

5 DISCUSSION

The obtained results show that the respondents are in

accordance with the most of the questions. The

majority of the questions were evaluated with three

or four points.

Figure 6: Histogram (UB).

Table 14: Top 3 questions with the highest and lowest

evaluation.

Highest Evaluation Lowest Evaluation

Construct Question Mode Avg Question Mode Avg

2.11.1

2.11.1.1

2.11.3

4,15

3,92

2.1.9

2.2.2

2.4.21

2,38

2,46

2,54

PEOU

2.4.1

2.11.1

2.11.1.1

4,08

4,15

2.1.9

2.4.21

2.4.22

2,38

2,54

2,61

2.11.1

2.11.1.1

2.11.3

4,15

3,92

2.1.9

2.3.2

2.3.3

2,38

2,31

2,38

2.10.1.2

2.10.1.4

2.11.3

3,77

3,69

3,92

2.1.9

2.2.2

2.3.2

2,38

2,46

2,31

To achieve the objectives originally proposed, a

questionnaire was considered based on the four

constructs (all areas) of TAM 3. The questionnaires

addressed all the components (system features)

exploited by the user in the ICU. The constructs with

the biggest acceptance degree were studied by

question and by each construct. For a better

understanding of the results (Table 14), a selection

was made on the three questions (Table 15)

associated to the higher valuation (best results) and

the three that presented the lowest valuation (worst

results). Table 16 shows that the greater acceptance

was the construct PEOU with an average of 3.45. At

the opposite side is the Using Behaviour with an

average of 3.23.

Table 15: ID and questions.

ID Question

2.1.9

Can help to mitigate situations of an excessive

workload?

2.10.1.2 Utility of GLASGOW CHART?

2.10.1.4

The graphics can help to a better understanding

of the real patient’s condition?

2.11.1 Utility of Information?

2.11.1.1

Utility of consulting information (hourly, daily,

continuous)?

2.11.3 Global evaluation of the vital signs?

2.2.2 Can access to information quickly?

2.3.2

Do you think that other nurses should use the

system as well?

2.3.3

Other professional colleagues think that you

should use the system?

2.4.1 Monitoring of the patient?

2.4.21 The Balance is done correctly?

2.4.22 Evaluation of Performance (speed)?

Table 16: Global Analysis for each construct.

Constructs Mode Average

PU 3 3,34

PEOU 4 3,45

BI 3 3,34

UB 3 3,23

For a best visualization, was also made a global

analysis on all the questions in the questionnaire

(91). The tables 17 and 18 present the three

questions with highest and lowest evaluations.

Table 17: Three questions with the highest evaluation.

Mode Avg

2.4.1 – Monitoring of Patient 4 4,08

2.11.1 – Utility of Information 4 4,15

2.11.1.1 – Utility of Consulting 4 4,15

Table 18: Three questions with the lowest evaluation.

Question Mode Avg

2.1.9 – Can help to mitigate situations of a

excessive workload?

3 2,38

2.2.2 – Can access to information quickl

2 2,46

2.3.2 – Do you think that other nurses shoul

use the system as well?

2 2,31

After an analysis of each construct, it was

performed a global analysis of all the responses

given by all the respondents. The answers presented

an average value of 3.40.

In general, the nurses are satisfied with the ease

of use of the technology. However, an obstacle is

limiting a wider acceptance of the system INTCare:

the operating speed of the user interface. This

PervasiveInformationSystemstoIntensiveCareMedicine-TechnologyAcceptanceModel

183

implies an upgrade of the workstations in terms of

RAM in the UCI. Another problem remains: the

most part of nurses have no time available to operate

with the information system.

6 CONCLUSIONS

The use of Technology Acceptance Model (TAM 3)

combined with Delphi method to evaluate the

acceptance by users, in order to understand their

perceptions and impact on the behaviour of the

system INTCare utility, is totally new. A set of

questionnaires based on the four constructs have

been answered by the nurses. In order to get a good

understand of the technology acceptance by the

users a set of analysis of the results (average, mode,

Kendall, better and worst features) were performed

having in consideration the TAM 3 methodology.

Certain limitations persist in the data access, due to

constant complaints from nurses regarding the speed

of system. The acceptance of the technology by the

nurses was very positive (between 3-4 points) for the

four constructs evaluated (Perceived Usefulness,

Perceived Ease of Use, Behavioural Intention and

Use Behaviour).

In the future, the results will be used to: improve

de system, mitigate some reported problems and add

some new features. Then, will be performed another

round of questionnaires, in order to understand if

there was some improved to the user at level of

TAM constructs.

ACKNOWLEDGEMENTS

This work is supported by FEDER through

Operational Program for Competitiveness Factors –

COMPETE and by national funds though FCT –

Fundação para a Ciência e Tecnologia in the scope

of the project: FCOMP-01-0124-FEDER-022674.

The authors would like to thank FCT

(Foundation of Science and Technology, Portugal)

for the financial support through the contract

PTDC/EIA/72819/ 2006 (INTCare) and PTDC/EEI-

SII/1302/2012 (INTCare II). The work of Filipe

Portela was supported by the grant

SFRH/BD/70156/2010 from FCT.

REFERENCES

Bolboaca, S. D., & Jantschi, L., (2006). Pearson versus

Spearman, Kendall's tau correlation analysis on

structure-activity relationships of biologic active

compounds. Leonardo Journal of Sciences, 5(9), 179-

200.

Chooprayoon, V., & Fung, C. C., (2010). TECTAM: An

Approach to Study Technology Acceptance Model

(TAM) in Gaining Knowledge on the Adoption and

Use of E-Commerce/E-Business Technology among

Small and Medium Enterprises in Thailand.

Gago, P., Santos, M. F., Silva, Á., Cortez, P., Neves, J., &

Gomes, L., (2006). INTCare: a knowledge discovery

based intelligent decision support system for intensive

care medicine. Journal of Decision Systems.

Hammer, Ø., Harper, D. A. T., & Ryan, P. D., (2001).

PAST-Palaeontological statistics. www. uv. es/~

pardomv/pe/2001_1/past/pastprog/past. pdf, acessado

em, 25(07), 2009.

Johns, R., (2010). Likert Items and Scales. Survey

Question Bank: Methods Fact Sheet, 1.

Orwat, C., Graefe, A., & Faulwasser, T., (2008). Towards

pervasive computing in health care - A literature

review. [10.1186/1472-6947-8-26]. BMC Medical

Informatics and Decision Making, 8(1), 26.

Portela, F., Santos, M., Vilas-Boas, M., Rua, F., Silva, Á.,

& Neves, J., (2010). Real-time Intelligent decision

support in intensive medicine. Paper presented at the

KMIS 2010- International Conference on Knowledge

Management and Information Sharing.

Portela, F., Santos, M. F., Silva, Á., Machado, J., &

Abelha, A., (2011). Enabling a Pervasive approach

for Intelligent Decision Support in Critical Health

Care. Paper presented at the HCist 2011 –

International Workshop on Health and Social Care

Information Systems and Technologies.

Saha, D., & Mukherjee, A., (2003). Pervasive computing:

a paradigm for the 21st century.

[10.1109/MC.2003.1185214]. IEEE Computer, 36(3),

25-31.

Santos, L. D. d., & Amaral, L. (2004). Estudos Delphi

com Q-Sort sobre a web: a sua utilização em sistemas

de informação.

Santos, M. F., Portela, F., Vilas-Boas, M., Machado, J.,

Abelha, A., & Neves, J., (2011). INTCARE - Multi-

agent approach for real-time Intelligent Decision

Support in Intensive Medicine. Paper presented at the

3rd International Conference on Agents and Artificial

Intelligence (ICAART), Rome, Italy.

Varshney, U., (2009). Pervasive Healthcare Computing:

EMR/EHR, Wireless and Health Monitoring:

Springer-Verlag New York Inc.

Venkatesh, V., & Bala, H., (2008). Technology acceptance

model 3 and a research agenda on interventions.

Decision Sciences, 39(2), 273-315.

Venkatesh, V., & Davis, F. D., (2000). A theoretical

extension of the technology acceptance model: Four

longitudinal field studies. Management science, 186-

204.

Zackiewicz, M., & Salles Filho, S., (2010). Technological

Foresight–Um instrumento para política científica e

tecnológica. Parcerias estratégicas, 6(10).

ICEIS2013-15thInternationalConferenceonEnterpriseInformationSystems

184