INTRODUCING A MOBILE SYSTEM FOR THE EARLY

DETECTION OF CARDIAC DISORDERS AS A PRECAUTION

FROM A CARDIOLOGISTS’ VIEW

Evaluation of a Survey

Onur von Burg, Marco Savini, Henrik Stormer and Andreas Meier

University of Friborg, Boulevard de Prolles 90, 1700 Fribourg, Switzerland

Keywords:

Cardiology m-health e-health mobile survey heart.

Abstract:

This paper illustrates the results of a survey of practicing cardiologists that were asked about various aspects

of a simpliﬁed telecardiology scenario using mobile devices. Such devices are becoming ubiquitous assets in

everybody’s life. Their application in a healthcare environment aims not only at supporting the patients over

traditional consultations but also maximizes the content of their health status information. The results of the

survey may help application developers to focus their efforts of applications in a similar setting.

1 EHEALTH AND MHEALTH

(Haas, 2006) deﬁnes eHealth as

...the application of Internet and other re-

lated technologies in the health care industry

to improve access, efﬁciency, effectiveness,

and quality of clinical and business processes

used by health care organizations, practition-

ers, patients, and consumers to improve the

health status of patients.

mHealth extends this notion by considering mo-

bile actors (patients, health professionals, ...) and

their environment. (Istepanian et al., 2006) simply de-

ﬁne it as ”emerging mobile communication and net-

work technology for health care”. Certain areas of

eHealth like monitoring over distance can be greatly

improved when considering ubiquitous mobile de-

vices. (del Pozo et al., 2006) deﬁnes that potential

needs should meet certain criteria so that:

• eHealth services should be provided at any time

and any place, depicting the need for efﬁcient,

ubiquitous and secure institutional care.

• Efﬁcient coordination tools should be provided to

all health professionals dealing with each single

patient, in order to allow the implementation of

patient centered care.

The recent advances in mobile technology like

higher bandwidths or powerful smartphones with ad-

vanced features allow the creation of new medical ap-

plications that support both patients and doctors.

A number of researchers have worked on the idea

of assigning mobile devices to patients. It is possible

to distinguish between the following three domains:

1. Mobile devices are used to help the patient by pro-

viding information.

2. Mobile devices are used to transmit physiological

parameters.

3. Mobile devices are used to alert patients or med-

ical professionals when certain physiological pa-

rameters become critical.

The dialogue-based monitoring system in

(Komninos and Stamou, 2006) aims at supporting

elderly people in their preferred environment and can

therefore be assigned to the ﬁrst domain.

The MOEBIUS project (Mobile extranet-based

integrated user services) that integrates doctors and

patients by submitting different physiological param-

eters (Fischer et al., 2006) can be considered as an ap-

plication of the second domain. Similarly, (Leimeis-

ter et al., 2005) describes the usage of mobile devices

in order to assist young cancer patients and concludes

that the usage of such a system has a number of ad-

vantages: higher compliance of appointments with

alerting functionality, higher data quality, less work

for the doctor to prepare the documentation as well as

less errors in the documentation. The use of context

von Burg O., Savini M., Stormer H. and Meier A. (2008).

INTRODUCING A MOBILE SYSTEM FOR THE EARLY DETECTION OF CARDIAC DISORDERS AS A PRECAUTION FROM A CARDIOLOGISTS’

VIEW - Evaluation of a Survey.

In Proceedings of the First International Conference on Health Informatics, pages 44-50

 SciTePress

in a mobile healthcare scenario using the information

groups Who, Where, When, Why and What has been

described in (Savini et al., 2007).

An example of the third domain collecting real-

time electrocardiogram signals including basic ar-

rhythmia detection with automatic alerting to a call

center is illustrated in (Liszka et al., 2004). Their

system architecture uses readily available commercial

off-the-shelf components.

1.1 Cardiology

1.1.1 Overview

Cardiovascular reasons represent the largest cause of

death in Switzerland with 37% of all deaths in 2004

(Junker, 2007). It is followed by cancer related deaths

with 26%. While the percentage of cardiovascular

related deaths of the 65 years and older increases

to 41.7%, it still contributes with 20% of all deaths

within the youngerage group of the 45 to 64 year olds.

Similar data exists in the EU with approximately 40%

of all deaths related to cardiovascular diseases (Euro-

stat, 2002, Chapter 8).

1.1.2 Cardiac Disorders

In (Guidant, 2007) the following general types of

heart diseases are identiﬁed:

Problems with the Heart’s Electrical System. If

the tissue that produces and sends electrical im-

pulses to the heart muscle is damaged, abnormal

heart beatings can occur in different variants:

Bradycardia (unusually slow or unsteady rhythm),

Tachycardia (more than 100 beats per minute),

Atrial ﬁbrillation (the upper chambers of the heart

beat between 300 and 600 times per minute) and

Sudden Cardiac Death (heart stops to beat and

pump suddenly and unexpected).

Problems with the Heart’s Arteries. These prob-

lems represent the most common heart disease

in western countries. Fatty plaque builds up in

the arteries and less blood ﬂows through the heart

muscle. This is illustrated in ﬁgure 1.

Problems with Arteries Outside the Heart. This

disease type also identiﬁes the building up of

fatty plaque, but outside the heart, so less blood

ﬂows through the body. If the blocked artery is

supplying the brain with blood, the patient may

experience a stroke in the worst case.

A thorough introduction into cardiology is given

in (Crawford et al., 2003).

Figure 1: Coronary heart disease (Know Heart Diseases,

2007).

1.1.3 ECG

An electrocardiogram can be deﬁned as a non-

invasive test to reﬂect underlying heart conditions by

measuring the electrical activity of the heart and pro-

ducing an output in the form of a continuous strip

graph. Early analysis and exploration of the heart’s

electrical system and its measuring techniques were

examined by Willem Einthoven in the early 1900’s

(Biel et al., 2001).

The basic approach is to put electrodes on the

body. The following three types of ECGs can be iden-

tiﬁed:

Resting. Taken within a couple of seconds.

Activity. Taken under physical activity.

Long-term. Usually taken during a 24 hour period.

The number of leads indicates the quality of the

measurement. 12 leads ensure that all parts of the

heart are recorded. It is generally accepted that ECG

measurements with 3 leads ensure credibility (Burger

and Osswald, 2003).

Various research has already been conducted re-

garding the transmission of ECG data. (Nollo et al.,

2000) describes the national project ”Territorial Tele-

Cardiology” which focuses on providing experts with

EGC data from medical ambulatory locations and pa-

tients homes. The average time of response is 1.5

days. The project demonstrated the feasibility and

utility of a telecardiology network that offers im-

proved access and quality in rural areas.

The design of a wearable device for ECG moni-

toring has been presented in (Led et al., 2004). The

paper describes the usage of Bluetooth technology to

transfer the ECG data to a mobile base unit.

Compression of ECG data is a topic that is very

important in a homecare scenario, due to the limited

bandwidth available over mobile networks and the

high amount of data which is generated. An overview

of various methods and research for the compression

of biosignals is given in (Hadjileontiadis, 2006).

Table 1: Return rates.

Region Infrastructure Size of Location

German French Hospital Practice > 100’000 > 30’000 < 30’000

35.7% 35.6% 61.5% 33.1% 29.2% 26.3% 45.4%

2 SURVEY RESULTS

The main motivation for the survey was to understand

how cardiologists think about a telemonitoringsystem

as an instrument for the detection of heart-related dis-

orders. The main goal hereby was to research various

components of how such a system should be built and

what focus it should have in order to receive a high

acceptance with the people that will eventually have

to introduce such a system to the patients.

2.1 Approach

The printed questionnaires were sent on the 13th of

March 2007 to 277 (out of totally 533) practicing car-

diologists in Switzerland. French and German ques-

tionnaires were used to address the corresponding lin-

guistic regions in Switzerland. The Italian region was

not considered in this survey. 100 questionnaires were

returned without any special follow ups, resulting in

a total return rate of 36.1%. Table 1 illustrates the

various return rates relative to various variables.

2.2 Structure of the Questionnaire

In order to design the questionnaire, some aspects

from (Berekhoven et al., 2006) were applied:

• Interesting subject matter.

• Questionnaire as short as possible.

• Composition as interesting as possible.

• Questions as logical and easy as possible (optical

as well).

Prior to the creation of the questionnaire an inter-

view was held with a practicing cardiologist who has

some background in the ﬁeld of telecardiology. The

result was a total of 12 questions that can be grouped

into two categories:

• Questions regarding ICT usage in cardiology in

general.

• Speciﬁc mobile telecardiology application related

questions.

A simpliﬁed process of a possible mobile telecar-

diology use case was illustrated on the ﬁrst page of

the questionnaire in order to give the respondents an

idea about the intention of the survey.

The box-and-whisker diagrams used throughout

this paper in order to illustrate the responses use a

ﬁlled circle as median of the response and the po-

sition of the dot within the box gives an indication

about the skew of the data. The dotted lines deﬁne the

whiskers which can be seen as the smallest or biggest

non-outlier observations. Eventual outliers are drawn

as empty circles.

2.3 ICT in Cardiology

Two questions address the use of ICT within cardi-

ology. In one question, the cardiologists were asked

about their opinion on how they judge the desire of

their patients to have innovative applications for treat-

ment and diagnosis. 28.3% of the cardiolists think

that their patients have a fairly strong desire for in-

novative applications. The correlations of these re-

sults with the demographic variables shows that the

geographical region or the size of the location of the

cardiologists does not inﬂuence the results. However,

there is a correlation to the type of infrastructure the

cardiologists operate in, indicating that cardiologists

working in hospitals evaluate the desire of patients for

new medical applications slightly higher than cardiol-

ogists in practices (see ﬁgure 2).

(1: strongly disagree, 5: strongly agree)

1 3 5

Hospital

1 3 5

Practice

Figure 2: ICT usage for patients in cardiology.

The second question regarding ICT addresses the

perspective of the cardiologists and asked if ICT is

used sufﬁciently in cardiology. The mean is very sim-

ilar to the former question. Compared to the infras-

tructure the cardiologists operate in, there is no dif-

ference between practices and hospitals, unlike to the

former question. The results for both questions are

illustrated in ﬁgure 3.

In summary, the followingstatements can be made

in regard to ICT usage in cardiology:

(1: strongly disagree, 5: strongly agree)

1 3 5

ICT for Cardiologists

1 3 5

ICT for Patients

Figure 3: ICT usage in cardiology.

• Patients have a certain desire for innovative appli-

cations in cardiology.

• Hospitalized patients have a higher desire for ad-

ditional ICT in cardiology than patients in prac-

tices.

• Cardiologists feel that today’s ICT usage in cardi-

ology is fairly sufﬁcient.

2.4 Motivation and Components in

Telecardiology

Questions regarding the components in a telecardiol-

ogy setting can be broken down into three main cat-

egories: The Patient category analyzes the abilities

and the reasons of patients to use such a system. The

questions in the category Patient-end System Compo-

nents emphasize on various scenarios and number of

leads that are necessary to measure an ECG. Further-

more, it investigates if the viewing of the ECG on

a mobile base unit makes sense. The ﬁnal category

Cardiologist-end System Components ad

2.4.1 Patient Ability

The cardiologists were asked to evaluate the ability

of their patients within certain age groups to record

and send their ECG data in a given scenario. The re-

sults shown in ﬁgure 4 are not surprising. Cardiolo-

gists judge that about half of their 60 to 70 years old

patients are able to record and send ECG data. The

conﬁdence is considerably higher for their younger

patients.

2.4.2 Reasons for Patients

Possible reasons for using such a telecardiology sys-

tem were investigated in a separate question. The total

results, independent of the infrastructure or geograph-

ical region, are summarized in table 2.

Not surprisingly, the number one motivation is

giving better access to patients living in rural areas.

Cardiologists living in hospitals consider this even

(1: none, 3: about half, 5: almost all)

1 3 5

30−45 45−60

60−70

1 3 5

70−100

Figure 4: Patient ability per age group.

Table 2: Possible reasons for patients.

Reason Agree

Cardiac patients who live in rural areas 58.6%

High quantity of elderly patients with

limited mobility

42.4%

Higher service in quality and treatment 39.4%

Growth of mobile users with easiness

to send ECG data

36.4%

Time gain for patients 32.3%

Reduction of costs 18.2%

more important (75%) than their colleagues in prac-

tices (55%). The next four reasons are all rated with

about the same importance. Interestingly, altough car-

diologists assess elderly patients as less able to use

such a mobile application (see ﬁgure 4), their limited

mobility is ranked as number two reason. Further-

more, the possible time gain for patients using such

a solution is judged positively by 50% of all cardi-

ologists in hospitals as an important reason, whereas

only 29% of practices see it like that.

2.4.3 System Components for Patients

This section of the questionnaire addresses various

component settings and scenarios on the patient side

for a mobile telecardiology application.

The ﬁrst question considers various usage scenar-

ios for patients and asked the cardiologists to evaluate

their view which is illustrated in ﬁgure 5.

Only one of the proposed scenarios, the perma-

nent transmission of ECG data for high risk patients,

received a relatively high response. Regarding the

periodic transmission of ECG data as a precaution,

the cardiologists in big cities valued it slightly higher

(mean 2.06) than their colleagues in towns (mean

1.79); this could be explained with the fact that car-

diologists in big cities tend to treat more patients for

routine check-ups.

(1: absolutely unnecessary, 5: absolutely necessary)

1 3 5

Analysis of ECG data during

physical activity

for competitive athletes

Periodic transmission as

precaution for Sudden Child

Death Syndrome

Periodic transmission for

adults to cardiologists as

a precaution

1 3 5

Permanent wearing of sen−

sors with connection to an

emergency center for

high rish cardiac patients

Figure 5: Usage scenarios.

The second question addresses the number of

EGC leads necessary for a mobile application in order

to reliably record the ECG of a patient. The distribu-

tion in ﬁgure 6 clearly illustrates two preferences: 3

leads and 12 leads.

Percent

1 lead 3 leads 6 leads 12 leads

Figure 6: Number of leads preference.

The evaluation of the effectiveness of diagnostics

using a mobile application for the three main types of

cardiac diseases is addressed in the third question and

its results illustrated in ﬁgure 7.

(1: not diagnosable, 5: very effective diagnosis)

1 2 3 4 5

Problems with

the arteries out−

side the heart

1 2 3 4 5

Problems with

the hearts’ elect−

rical system

1 2 3 4 5

Problems with

the hearts’

arteries

Figure 7: Evaluation of disease types.

Problems with the arteries outside the heart do not

represent situations that can be diagnosed using ECG

sensor leads. In relation to the number of leads, pa-

tients that haveproblems with the hearts’ arteries typi-

cally will use 12 leads and such applications will have

to consider the larger data sets required. For patients

that use such a mobile application as precaution for

the heart’s electrical system, a 3 lead ECG sensor ap-

pears to be more realistic.

The ﬁnal question regarding the system compo-

nents on the client side investigates the need of the

patient to view his own ECG data on the mobile base

unit. The results in ﬁgure 8 show that cardiologists do

not consider such a functionality very desirable for

their patients. This might be due to the fact that ECG

diagrams can be quite complex to read and should

thus only be analyzed by experienced professionals.

(1: no necessity, 5: strong necessity)

1 2 3 4 5

Figure 8: Patient desire to view ECG on MBU.

2.4.4 Components in Telecardiology for

Cardiologists

The ﬁnal part of this section considers the various

components of a mobile telecardiology solution for

the cardiologists themselves. It consists of two ques-

tions: The ﬁrst question addresses the ICT knowledge

of the cardiologists based on possible components or

technologies that could be used for such a solution

and is illustrated in ﬁgure 9.

Internet applications, mobile communication, in-

formation system and PDAs rank highest in the list

and except for technical terms many possible compo-

nents and technologies seem to be known reasonably

well. Cardiologists in practices tend to have similar

or slightly higher knowledge in regard to Internet Ap-

plications, as ﬁgure 10 illustrates. The same is true

for mobile technologies and other ﬁelds.

This explains also the results of the next question:

The cardiologists were asked about their preferences

how they would like to receive the ECG data of the

patients for the various scenarios already interrogated

in ﬁgure 7. A summary of the preferences is ranked

in table 3.

However, the results in ﬁgure 11 make it obvi-

ous that the preferred end system heavily relies on the

kind of disease type that is handled with the solution.

Number of Responses

Application

E−Mail

Smartphone

Webbased

Analysis of ECG data during physical activity

for competitive athletes

Application

E−Mail

Smartphone

Webbased

Periodic transmission as precaution for

Sudden Child Death Syndrome

Periodic transmission for adults to cardio−

logists as a precaution

Permanent wearing of sensors connecting to an

emergency center for high risk patients

Figure 11: Preferred end systems per disease type.

(1: never heard of, 3: user knowledge, 5: extensive knowledge)

1 3 5

Bluetooth GPRS

1 3 5

GSM

Information

Systems

Internet

Applications

Mobile

Communications

PDA

1 3 5

Smart Phones UMTS

Figure 9: ICT knowledge of cardiologists.

(1: never heard of, 3: user knowledge, 5: extensive knowledge)

1 3 5

Hospital

1 3 5

Practice

Figure 10: Knowledge about internet applications.

Web-based solutions seem to be a good choice in

most cases. However, smart phones get a very high

rating when the application handles high risk patients

Table 3: Ranked summary of preferred end systems.

System Preference

Web-based solution 38.7%

Smart Phone 24.3%

E-Mail with attachment 21.2%

Dedicated application 15.8%

in an emergency. In such a situation, an e-mail that

may take from minutes up to hours to arrive to the

cardiologist, is not acceptable, even if the emergency

is handled by a dedicated center. However, e-mails

seems to be perfectly acceptable if the matter is about

observing a patient over a longer period of time as a

precaution.

3 CONCLUSIONS AND

OUTLOOK

The work in this paper focuses on the actual needs of

a mobile telecardiology application with the results

obtained from the survey. The conducted survey al-

lows several conclusions for the implementation of a

mobile application in cardiology:

• Except for the oldest age group (70-100), cardiol-

ogists agree that their patients are able to use such

a system.

• Cardiologists do not seem enthusiastic about the

proposed usage scenarios. However, a combina-

tion of factors, such as high risk patients in rural

areas, may strongly contribute to a higher accep-

tance.

• Addressing problems with the hearts’ electrical

system seems to be a widely accepted disease type

for such a mobile application.

• There does not seem to be a strong desire for the

patients to view their own EGC.

• The ICT knowledge of cardiologists and their

preferences imply that, except for emergency sit-

uations, an internet application would be a good

choice for their end system.

Security issues have explicitly been factored out

of the questionnaire in order not to overwhelm the

cardiologists with technical details. However, such

issues are a pivotal part of every health related ap-

plication and contribute highly to the acceptance and

credibility of a system.

3.1 Outlook

In this paper we illustrated the perspective of cardiol-

ogists regarding a mobile telecardiology application

scenario for patients. We would like to extend our

research in the following areas:

• A survey on the patients may give insights about

what elements to consider on the client side. It

would be interesting to analyze the technologi-

cal knowledge split by age group and the willing-

ness of patients to use such a system when offered

some well-deﬁned beneﬁts.

• The usage of standard equipment greatly con-

tributes to a seamless rollout to the public. We

would like to investigate how existing technolo-

gies can be integrated in order to provide a so-

lution that aids the patients and cardiologists and

that is also cost-effective.

• Qualitative interviews with cardiologists might

give us a better idea about possible usage scenar-

ios.

REFERENCES

Berekhoven, L., Eckert, W., and Ellenrieder, P. (2006).

Marktforschung – Methodische Grundlagen und prak-

tische Anwendung. Gabler Press.

Biel, L., Pettersson, O., Philipson, L., and Wide, P. (2001).

ECG Analysis – A new approach in Human Identi-

ﬁcation. IEEE Transactions on Instrumentation and

Measurement, 50:808–812.

Burger, P. and Osswald, S. (2003). Das EKG in der

Telemedizin. Swiss Medical Forum, 3:661–643.

Crawford, M. H., DiMarco, J. P., and Paulus, W. J., editors

(2003). Cardiology. Mosby, 2 edition.

del Pozo, F., de Toledo, P., Jimenez, S., Hernando, M. E.,

and Gomez, E. J. (2006). Chronic Patient’s Manage-

ment: The COPD example. In M-Health – Emerging

Mobile Health Systems, pages 575–585. Springer.

Eurostat (2002). Health statistics – Atlas on mortality in the

European Union. European Commission.

Fischer, H.-R., Reichlin, S., Gutzwiller, J.-P., Dyson, A.,

and Beglinger, C. (2006). Telemedicine as a new pos-

sibility to improve health care delivery. In M-Health

– Emerging Mobile Health Systems, pages 203–218.

Springer.

Guidant (2007). Heart Disease and Cardiovascular Disease

Therapies. http://www.guidant.com. accessed on 5th

June 2007.

Haas, P. (2006). eHealth ver¨andert das Gesundheitswesen

– Grundlagen, Anwendungen, Konsequenzen. HMD

Praxis der Wirtschaftsinformatik, 251:6–19.

Hadjileontiadis, L. J. (2006). Biosignals and Compression

Standards. In M-Health – Emerging Mobile Health

Systems, pages 277–292. Springer.

Istepanian, R. S., Laxminarayan, S., and Pattichis, C. S.

(2006). Ubiquitous m-health systems and the con-

vergence towards 4G mobile technologies. In M-

Health – Emerging Mobile Health Systems, pages 3–

14. Springer.

Junker, C. (2007). Todesursachenstatistik – Ursachen der

Sterblichkeit 2003 und 2004. Bundesamt f¨ur Statistik.

Know Heart Diseases (2007). Know Heart Diseases.

http://www.know-heart-diseases.com. accessed on 5th

June 2007.

Komninos, A. and Stamou, S. (2006). HealthPal: An Intel-

ligent Personal Medical Assistant for Supporting the

Self-Monitoring of Healthcare in the Ageing Society.

In Proceedings of UbiHealth 2006.

Led, S., Fernandez, J., and Serrano, L. (2004). Design of a

Wearable Device for ECG Continuous Monitoring Us-

ing Wireless Technology. In Proceedings of the IEEE

EMBS, pages 3318–3321.

Leimeister, J. M., Krcmar, H., Horsch, A., and Kuhn, K.

(2005). Mobile IT-Systeme im Gesundheitswesen,

mobile Systeme f¨ur Patienten. HMD – Praxis der

Wirtschaftsinformatik, 244:74–85.

Liszka, K. J., Mackin, M. A., Lichter, M. J., York, D., Pillai,

D., and Rosenbaum, D. S. (2004). Keeping a Beat on

the Heart. IEEE Pervasive Computing, 3(4):42–49.

Nollo, G., Nones, M., Malossi, C., Conforti, F., and Bato-

cletti, B. (2000). Standard communication protocol

for ECG exchange in tele-cardiology network for ru-

ral areas: three years project experience. Computers

in Cardiology, 27:663–666.

Savini, M., Stormer, H., and Meier, A. (2007). Integrat-

ing Context Information in a Mobile Environment us-

ing the eSana Framework. In Proceedings of the

ECEH’07, pages 131–142.