The User Interface of a Medical First Aid Application

A Close-to-Realistic Usability Study with the Smartphone Application “Defi Now!”

Karin Harbusch and Janine Paschke

Department of Computer Science, University of Koblenz-Landau, Koblenz, Germany

Keywords: First aid, Cardiopulmonary Resuscitation, CPR, Automatic External Defibrillator, AED, Bystander Effect,

Smartphone App, Usability Study, Induced Emotion, Software-ergonomics, Human-computer Interaction.

Abstract: Several smartphone applications are available to support the delivery of medical first aid, in particular car-

diopulmonary resuscitation (CPR), by untrained people. These apps support passers-by witnessing an emer-

gency situation, or enable brushing up basic knowledge of first aid. Moreover, they inform about publicly

available automatic external defibrillators (AEDs) and allow users to add the location of new devices. In

this paper, we describe features of various first aid apps currently on the market. In order to find out whether

or not these apps can really be helpful in emergency situations, we conducted a usability study with 74 par-

ticipants who used the “Defi Now!” app. In order to simulate “realistic” conditions where the user is agitat-

ed—as is the case when rescuing a person suffering from cardiac arrest—we induced fear by a psychologi-

cally recommended method. Based on data from a questionnaire and video recordings, we discuss strong

and weak points of “Defi Now!”. The app was judged to be very helpful for medically untrained people.

Nonetheless, our observations suggest several improvements to the user interface.

1 INTRODUCTION

According to the Federal German Health Infor-

mation System (http://www.gbe-bund.de), sudden

cardiac arrest due to ischemic heart disease is the

most common cause of death in most Western coun-

tries, and a major cause of hospital admissions. In

the U.S., the number of deaths due to Sudden Cardi-

ac Death exceeds 250,000 annually.

In case of cardiac arrest, immediate help is es-

sential and defibrillation is mostly inevitable; to this

purpose, automatic external defibrillators (AEDs)

are currently provided at many public places. Aver-

age survival rates of 30% for defibrillation by first

responders within 3 to 5 minutes are reported by

http://www.heart.org/HEARTORG/. The chance of

survival decreases by 10% per minute when no help

is provided (e.g., Cummins et al., 1985). Hence,

everybody who owns a feature phone or smartphone

is obliged to call for help on site. Moreover, passers-

by should administer first aid, especially cardiopul-

monary resuscitation (CPR)—a procedure helping to

restore blood circulation and to preserve intact brain

function—, when a person is suffering from cardiac

arrest or shows no vital signs. CPR (either per-

formed as compression-only without mouth-to-

mouth or with mouth-to-mouth respiration) is a

manual treatment, indicated until further treatment,

such as defibrillation, is available.

But in such a precarious situation, the bystander-

effect (Darley and Latané, 1968) is more likely to be

observed than immediate help, i.e. the phenomenon

that, in case of an emergency, individuals are less

willing to offer help the more bystanders are present.

This diffusion of responsibility results in waiting for

someone else to take command. Furthermore, by-

standers tend to belittle the existing emergency since

no one has intervened so far. Thus, they judge that

no emergency is at hand and help is not necessary.

The most common cause for the denial of assistance,

however, might be the fear to fail or to be ridiculed.

This fear discourages many people to help, especial-

ly when they are unable to judge the gravity of the

situation (e.g., Jörg, 2012). To what extent can a

smartphone application (app) help passers-by not to

become bystanders but rescuers?

In the next section, we summarize desirable fea-

tures of existing first aid smartphone applications.

In order to find out whether or not these apps can

really be helpful in an emergency case, we investi-

gated the usability of the “Defi Now!” app with 74

participants. In order to simulate the stress that is

Harbusch K. and Paschke J..

The User Interface of a Medical First Aid Application - A Close-to-Realistic Usability Study with the Smartphone Application “Deﬁ Now!”.

DOI: 10.5220/0004441100910098

In Proceedings of the 15th International Conference on Enterprise Information Systems (ICEIS-2013), pages 91-98

ISBN: 978-989-8565-61-7

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

typically experienced by untrained individuals who

are helping a person suffering from cardiac arrest,

we applied induced agitation. In Section 3, we pre-

sent the results of our study. Basically, the app was

judged to be very helpful. However, our observa-

tions during the experimental sessions suggest sev-

eral software-ergonomic improvements in support of

the rescuers (cf. Section 4). In Section 5, we draw

conclusions and address future work.

2 FIRST AID APPS

A search for “first aid” applications in the markets

Google play and ITunes of the two main operating

systems for smartphones yields more than 1,000

applications. However, this number includes enter-

tainment applications such as a gamers’ guide for

“World of Warcraft”. Here, we list six apps that

focus on first aid in case of cardiac arrest and are

available in Austria, France, Germany, Switzerland,

or the United States:

 Defi Graz (www.madison.at (Austria); interface

only in German);

 Arrêt Cardiaque 2.0 (www.associationrmcbfm. fr

(France); multilingual interface);

 AED Locator (www.aedlocator.org (Germany); in

German);

 Defi Now! (definow.org (Germany); multilingual

interface);

 Herzsicher (www.herzsicher.ch (Switzerland); in

German);

 PulsePoint (pulsepoint.org (United States); in

English).

First, we outline the main features of “Defi

Now!” (Lange, 2011), the system we evaluated in

our study (Section 3). Then, we complement the list

of desired features with best practices from the other

apps, and recommend improvements that can yield

an optimal interface for a cardiac arrest first aid app

(Section 4).

The interface language of “Defi Now!” is auto-

matically adjusted to the user’s phone settings. For

example, if the user has chosen English as preferred

language, the app appears in English as well. When

opened, the app displays three options:

1. Emergency call;

2. First aid measures;

3. Find an AED.

Pressing the button “Emergency Call” automati-

cally activates the national emergency phone num-

ber adapted to the user’s whereabouts. However, the

button does not immediately dial the emergency

number because the rescuer has to confirm first by

pressing “Yes” in response to the question whether

they actually want to place the emergency call—in

order to avoid hoax calls.

While placing the call, the current address is dis-

played to the user so that foreigners (and nervous

rescuers) can easily provide the location’s precise

address (in Germany, GPS data are not automatical-

ly sent).

By pressing the button “First Aid Measures”,

first aid procedures are displayed in the form of

diagrams (Figure 1), separately for patients showing

vital signs (right branch from topmost node) and

patients showing no vital signs (left branch).

Figure 1: “Defi Now!” guideline for administering first

aid.

If the patient shows no vital signs, the rescuer re-

ceives information on how to perform cardiopulmo-

nary resuscitation (CPR). A metronome provides

audio support for a resuscitation beat of 100 times

per minute, with an option to administer mouth-to-

mouth/nose respiration. The user selects the desired

action by scrolling to the bottom of the page, where

two options are presented (see Figure 2): pressing a

text button, which activates the standard rhythm of

100 beats per minute, or pressing a settings wheel.

The latter option enables the user to choose the arti-

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

ficial respiration option, which activates a spoken

command telling the rescuer when to breath (two

times breathing after every sequence of 30 beats).

Figure 2: Select the metronome beat and the setting for

artificial respiration.

Activating the third item of the opening screen by

pressing the button “Find an AED” enables the

search for AED locations on an interactive map.

Four types of icons may appear (cf. the pins in the

center of the left panel of Figure 6):

1. A red cross on white background indicates a

place where medical staff is available;

2. Green lightning icons denote verified AEDs;

3. Grey lightning icons stand for still unconfirmed

AEDs;

4. Blue lightning icons refer to AED locations re-

trieved by the database “AED-Kataster”.

In order to update this database, the user can add

a new AED device (not discussed here due to

space limitations).

Now, we sum up desirable features present in the

abovementioned apps but not in “Defi Now!”. Sev-

eral other systems (e.g. “Herzsicher” and “Pulse-

Point”) provide a tab(ulator)-based interface (see

Figure 3), where all available functions are perma-

nently visible/selectable. This dialogue style avoids

error-prone search for, and navigation to, the desired

menu item. Therefore, in our study, we evaluate how

easy it is for “Defi Now!” users to deal with non-

tabulator-based navigation.

“Defi Graz” and “Arrêt Cardiaque 2.0” support

first aid in a more graphical manner (see Figure 4)

that might be easier to follow than a more textual

instruction such as in “Defi Now!”. Therefore, in our

study, we asked the participants how comprehensi-

ble they found the instructions given in the dia-

grams.

All non-German systems allow for easier emer-

gency call functions due to automatic location of the

smartphone.

‘Map’ ‘AEDs’ ‘Help’ ‘Add’ ‘More’

Figure 3: Top-Level Choice Menu in “Herzsicher” Exem-

plifying a Tabulator-based Interface.

Figure 4: Upper panel: Graphically supported first aid

instructions in “Defi Graz”; Lower panel: CPR details in

“Arrêt Cardiaque 2.0”.

“Arrêt Cardiaque 2.0” provides social-network

integration. The app may be connected with the

user’s Facebook and/or Twitter account to share the

AED they registered with their network. The inte-

gration of social media might serve as a fast way of

distributing information about AEDs to a great num-

ber of users. Hence it might promote public aware-

ness of this important information. The more people

are aware of such critical issues, the greater the

likelihood they also want to learn how to act in an

emergency instead of functioning only as bystand-

ers.

In “Arrêt Cardiaque 2.0” and “PulsePoint”, users

can register as first responders. If a cardiac arrest

occurs in the vicinity of a first responder, this person

is sent a notification in case of a cardiac arrest near-

TheUserInterfaceofaMedicalFirstAidApplication-AClose-to-RealisticUsabilityStudywiththeSmartphone

ApplicationDefiNow

by. For example if a cardiac arrest occurs in a shop-

ping center, it might be possible that a subscriber of

such an emergency notification may be in the same

building or in the neighborhood. In “PulsePoint”, the

thus alarmed rescuer receives a route description to

the location of the patient.

“PulsePoint” provides additional functionality

for emergency situations other than cardiac arrest.

Via a tab labeled “Incidents”, the user can ask for an

overview of all fire department emergencies occur-

ring at the current moment. A “Photos” tab provides

a photo gallery of major incidents from firedepart-

ment.org. Users may scroll through the gallery and

select photos they want to view. A “Radio” button

enables a streaming radio feed from the dispatch

center of the fire department.

3 THE USABILITY STUDY

WITH DEFI NOW!

Some of the features we outlined above (e.g.,

graphics) are obviously very suitable to make a first

aid app supportive for a rescuer. However, more

detailed insights into unexpected problems require a

usability study (see, e.g., Nielsen, 1994).

3.1 Method

The participants were tested in individual sessions.

They were seated at a desk, with an iPhone 3GS

(iOS 6.0.1) with the preinstalled “Defi Now!” app in

front of them, and carried out the tasks listed in our

questionnaire (see Section 3.2). Their actions were

recorded via a camcorder (Plawa DV-4 SD-

Camcorder; image resolution 4.0 MP), which was

focused on the iPhone’s display, thus ensuring ano-

nymity of the participants. The videos were used to

reconstruct the handling of the application, and to

survey the errors made during task execution.

We tested 74 participants, 35 female and 39 are

male; for their age distribution, see Figure 5. Nine of

them had been, or still were, students of computer

science at our campus, where we had run a pilot

study in an observation lab. Another eight partici-

pants were doctors or qualified medical employees.

The other 57 participants had different occupations,

such as fitness trainer, banker or steel mill worker.

Four participants had experienced a heart attack

themselves before; one of them even had suffered

from a sudden cardiac arrest and had to be resusci-

tated. All but the students were recruited as patients

in the waiting room of a doctor’s office or as cus-

tomers of a fitness center where we had reserved a

separate room to perform the experiment.

Figure 5: Age and gender distribution of the participants.

When witnessing an emergency situation such as

cardiac arrest, bystanders and rescuers usually show

signs of emotional stress, for example trembling

hands. In order to simulate this stress, we applied a

psychological technique to induce agitation (see,

e.g., Otto, 2000). Watching a short sequence of a

movie can elicit high levels of agitation in test par-

ticipants due to the high attention and the intensive

experience tied to watching a film clip (e.g., Gross

and Levenson, 1995); (Hewig et al., 2005); (Schlei-

cher, 2009). For our experiment, we selected a film

clip from the movie “The Shining” (1980), showing

a playing child who heads towards a closed hotel

room door. The viewer gets the feeling that some-

thing terrifying is lurking behind the closed door

(see also Hewig et al., 2005 on this fragment).

In order to study the effect of agitation, a random

selection of 54 of our participants watched the film

clip before using the app. The remaining 20 partici-

pants functioned as a control group and did not view

the film clip. Age, gender, ethnic background, etc.,

distributions in the two groups were comparable.

3.2 Questionnaire

The initial section of our questionnaire consisted of

questions concerning factors known to influence

precision of task execution (cf. Schleicher, 2009).

The participants had to state their personal condition

with respect of discomforts (e.g., high blood pres-

sure and/or use of medication). Moreover, they la-

beled their emotional state (Happiness, Grief, Fear,

Disgust, or No Emotion). They could also add a label

for the emotion they experienced if none of the pre-

defined ones matched. Additionally, they stated their

personal condition as good, relaxed, nervous, anx-

ious, tired, or ill-tempered.

The next part of the questionnaire consisted of

nine tasks, covering all features of the application

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

“Defi Now!” to be carried out during the test. After

each task, the participants had to judge the presenta-

tion of a given task by means of a five-point rating-

scale, ranging from 1=”not understandable at all”

to 5=”completely understandable” (cf. Rohrmann,

1978). Moreover, they could spontaneously mention

what they (dis-)liked about the task. For each of the

individual tasks we measured the time it took to

perform them.

The last part of the questionnaire consisted of

various general questions to the participant: Whether

they owned a smartphone themselves; whether they

find spoken instructions in the app helpful, especial-

ly with respect to administering first aid. Further-

more, they were asked whether/when they attended

their last first aid course, and whether they judged a

first aid app useful in case of emergencies such as

cardiac arrest.

3.3 Results

As regards the influence of stress, we could not

observe any significant effect on the execution of the

“Defi Now!” tasks. This is in spite of the fact that

the film clip induced emotional states comparable to

those reported by Schleicher (2009). In our study,

17% indicated they did not feel any emotion while

watching the clip. However, the intended emotion

was evoked in 70 % of the participants. We found no

significant differences between the student group,

the medically trained personal and the other partici-

pants. They all had similar problems with the same

features of the user interface.

Let us first describe the answers to the questions

about personal factors with influence on precision of

task execution. With respect to their physical condi-

tion, 84% declared they did not suffer from any

health problems. From the 16% who mentioned

health problems, one participant used tranquillizers,

two had a common cold, five are high-blood-

pressure patients and three had problems with their

cervical spine. Analysis of the video data and the

participants’ behavior, lead to the conclusion that

some discomforts as well as specific medication

could affect a person’s performance. The participant

who used tranquillizers, for example, seemed dis-

tracted and agitated. Those who suffered from hy-

pertension did not show signs of nervousness. How-

ever, the physical condition of a participant had no

influence on their emotional state.

With respect to their current emotional state,

59% stated they felt good, 15% stated they were

relaxed and 14% were nervous. Five participants

(7%) declared they were tired and another two felt

anxious (3%). None of the participants answered

they were ill-tempered.

Now we sum up the observations recorded dur-

ing performing specific task with the app. (Due to

space, we focus here on first aid and finding an

AED). When the test participants were asked to open

the function “First Aid Measures” and to review the

diagrams with first aid procedures, 57% stated that

they knew which first aid step to perform next, at all

times. Only 12% felt uncertain about possible next

steps, 19% judged the diagrams as unsuitable and

wanted more detailed instructions. On the other

hand, 64% declared the diagrams suitable for learn-

ing, since they immediately understood the proper

sequence of first aid procedures to be carried out in

case of an emergency.

When asked whether they saw the circular sym-

bol in the diagram (cf. the lower right corner of the

box in the middle of the diagram in Figure 1), 64%

of the participants denied. 21% searched for the

mentioned symbol on request and found it after an

average of 7.6 seconds. Merely 16% of the partici-

pants had detected the circular symbol spontaneous-

ly. (Notice that the circle indicates the need to exe-

cute the corresponding measure repeatedly; without

repetition, CPR is ineffective.) Twelve participants

stated here that they would appreciate spoken state-

ments or a combination of speech and graphics/text.

Eight participants criticized the wording used in the

diagrams. The medical term “Compressions of the

thorax” was often judged as too technical. Seven

participants stated they had rather seen more

graphics and less text. One of them remarked that

“assembly instructions guide primarily via graphics

and not via text. If they would consist mainly of text,

people would probably not read it.”

Two of the participants who were doctors, com-

mented that the diagrams did not indicate that the

movement of the patient’s head should be a reclina-

tion, not a rotation. (If this movement is not per-

formed correctly, respiration cannot be performed

effectively).

Nevertheless, the most salient point of criticism

was the fact that the user has to exit the first aid

guidelines when placing an emergency call. Six

participants voiced the idea that the feature of plac-

ing an emergency phone call should be integrated

directly into the diagram of first aid measures, pos-

sibly visualized as a button at the onset of these

guidelines.

The App “Defi Now!” contains a metronome for

the user to set CPR rhythm (number of thorax com-

pression per minute) and CPR mode (with or with-

TheUserInterfaceofaMedicalFirstAidApplication-AClose-to-RealisticUsabilityStudywiththeSmartphone

ApplicationDefiNow

out artificial respiration). During this task, the users

were prompted to activate the metronome’s acoustic

signal supporting the pace of thorax compressions,

and then to stop the signal. This task proved to be

error-prone. Only 32% of the participants recognized

the trigger of the metronome immediately when they

scrolled to the bottom of the diagram (cf. Figure 2).

41% had more difficulties and pressed the button

predefining the resuscitation beat only after an aver-

age of 6.3 seconds. 5% initiated the beat after more

than 10 seconds and 23% saw the trigger only after

instruction. Overall, 58% expressed the desire for

more directions given by the application itself. 47%

suggested that the trigger should be placed at a dif-

ferent place, for example directly next to the

graphics depicting the CPR procedure.

If not stopped manually, the metronome keeps

producing the beats. When asked to stop the acoustic

signal, 47% were able to fulfill this task within 3.2

seconds. Nevertheless 41% required 5 seconds or

longer to pause the metronome. 65% criticized the

button label shown while the metronome was active.

While the beats go on, no label referring to stopping

it is visible; instead, in case of artificial respiration,

the button shows the number of beats passed by

since the last breathing.

Several participants voiced misgivings regarding

controllability, since it took them quite some time to

find the trigger setting the resuscitation beat and

then to stop it again. Some participants criticized

that a user probably would not even expect a metro-

nome to set the CPR rhythm, if they had not been

made aware of the availability of this feature. Ten

participants suggested highlighting the metronome

trigger, for example by labeling it more clearly,

giving it a different color or adjusting the button

size.

One of the participants’ tasks was to change the

settings for CPR execution; that is, to indicate

whether the 100 beats per second should be inter-

rupted, after every 30 beats, by the spoken command

to apply mouth-to-mouth respiration (two breath-

ings: 30:2). 47% declared they immediately recog-

nized the button referring to the setting of resuscita-

tion beats (cf. Figure 2). 53% experienced difficul-

ties fulfilling the task of changing the CPR mode

(with or without artificial respiration). 26% of the

participants held the opinion that this facility for

alteration was not relevant, whereas 74% found it

important. Six participants criticized that the settings

button was insufficiently salient and desired a more

obvious labeling, for example a written identifica-

tion of the function of the button.

Three participants (medics) remarked that the du-

ration of time scheduled for mouth-to-mouth respira-

tion, which is set to three seconds, might not suffice

because the rescuer has to change body position

when starting artificial respiration. Several partici-

pants (medical personnel and laymen) remarked that

the option to alter CPR mode might in fact be ob-

structive and counterproductive. They argued that a

user might be overchallenged by the options of car-

diac-only resuscitation vs. CPR with mouth-to

mouth respiration. For them, a precise instruction

preinstalled in the application would be preferable.

One user opposed to the fact that the metronome

stops as soon as the user quits the menu section

“First Aid Measures” and, for example, returns to

the map searching for an AED. The rescuer is unable

to adjust CPR to the pace preset by the metronome

while seeking an AED. They might also not be able

to keep pace with the metronome rhythm, which

might entail severe consequences, as stated before.

(Other passers-by might happen to have the “Defi

Now!” application installed on their smartphones,

and activate the beat there.)

To sum up the findings regarding the second top-

level menu item, the task of setting the CPR mode

brought several problems to light concerning opera-

bility. Several participants were not satisfied with

the implementation of features such as the labels on

the buttons referring to the starting/stopping of the

resuscitation beat and CPR mode. They evaluated

these features as being not self-explanatory and

expected clearer labels for better usability. Moreover

they were not satisfied with the controllability since

it took them quite some time to find the trigger start-

ing the metronome settings—in about 70% of the

cases it took more than 6 seconds to get the acoustic

signal started. Many participants said that in case of

an emergency, they would waste time searching for

those specific functions, and remarked that these

should be implemented in a more eye-catching and

better comprehensible manner.

The test of the third top-level menu item “Find

an AED” unveiled several unfulfilled user expecta-

tion as regards the user interface. The app shows

either a tabular view or an interactive map, depend-

ing on which icon in the upper panel is selected (in

the upper panel of Figure 6, see the highlighted

buttons serving to switch between the two option).

41% of the participants stated they expected to find

this function within the menu section concerned with

adding a new AED location; 54% expected the tabu-

lar view under “Find an AED”. Seven participants

commented on the tabular-view button. In their

opinion, a written labeling would have marked the

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

function clearer. Three participants voiced the idea

that the tabular view might be displayed before re-

vealing the map. This might facilitate the under-

standing of which AED location is the nearest, as

indicated by the distance value.

Figure 6: Switching map to list presentation of AEDs.

With respect to overall navigation, the app some-

times violated user expectations; e.g. the “Back”

button occurs in the lower panel although it is usual-

ly located in the upper left corner (in iPhones).

Unclear icons in unexpected positions increase oper-

ating time. An example is the “Information” button

for the types of AEDs. Although the icon “i” in a

circle is well chosen, its small size, its grey back-

ground color and its position in the lower right cor-

ner (see the left screen shot of Figure 6) made it

unfindable to nearly all participants.

Due to space, we only address two questions

from section 3 of the questionnaire here. 71% of the

participants agreed that “Defi Now!” is appropriate

when administering first aid in an emergency. How-

ever, 7% stated that the app might distract inexperi-

enced users who are unfamiliar with the app. Thus,

eliciting interest in first aid procedures without the

stress elicited by the emergency situation is an im-

portant issue.

Familiarity with first aid procedures did not cor-

relate with execution time of the tasks—opposite to

what one might have expected; according to the

questionnaire, 2 test subjects had never taken any

first aid course; 25 had followed a course less than 3

years ago; 11 less than 10 years ago; and, 35 partici-

pants more than 10 years ago.

4 OPTIMIZATING

THE USER INTERFACE

Based on our user study with “Defi Now!” and best

practice in other smartphone apps, we recommend

some easily attainable improvements for a first aid

user interface and illustrate them with mock-ups.

In view of German law, the installation dialog

could ask the user, as soon as the system has been

loaded, for permission to localize the smartphone;

this would provide for automatic emergency calls (of

course, after confirmation by the user). This feature

would also enable automatic route descriptions to

the next AED automatically, as in PulsePoint. The

app “Arrêt Cardiaque 2.0” suggests another very

desirable feature. “Defi Now!” should be embedded

in social network environments so that knowledge

about its availability can spread faster and wider.

In order to highlight all available functions, we

propose a three-part start menu (see Figure 7) where

the “Emergency mode” button is pre-selected as

default in order not to waste time selecting it under

time pressure. The second item can be an “Explora-

tion mode”, inviting smartphone owners to try out

and to learn about first aid and nearby AED loca-

tions in their spare time. The possibility of adding

new AEDs can become more prominent when it is

listed in the start-up menu (as third item).

Figure 7: Mock-up of the start menu.

Given the observed confusion when navigating, we

propose a tab-based surface (Figure 8) as used in

several other first aid apps (Section 2). Red color

indicates which tab is currently active. Moreover,

the emergency call tab (which in Germany is 112—

but should be adapted automatically to the user’s

current whereabouts) could become grey after being

used once.

The way first aid measures are presented could

receive better graphical support (as in Figure 4) and

be underpinned with spoken instructions (to be acti-

vated, for instance, by touching an image in the

overview). This issue should be worked out in col-

laboration with medical experts in order to avoid

TheUserInterfaceofaMedicalFirstAidApplication-AClose-to-RealisticUsabilityStudywiththeSmartphone

ApplicationDefiNow

ambiguity in the provided images and to select the

best CPR mode (e.g., based on survival rates report-

ed in a study by Bohm et al., 2007). Crucially, the

CPR instructions should be visible permanently

together with a nearby change-mode button. Alterna-

tively, the two CPR modes could be directly associ-

ated with the CPR box in the diagrams, recognizable

as push buttons (see the grey areas with loudspeaker

symbols in Figure 8). Switching audio on and off

should be possible in a way familiar to many pro-

spective users.

Figure 8: Mock-up illustrating always-visible tab-

navigation and improved first-aid diagrams.

5 CONCLUSIONS

We have outlined the functionality provided by

various apps aiming to support the delivery of first

aid to persons suffering from cardiac arrest, in cir-

cumstances where immediate help is essential and

defibrillation is virtually inevitable. We reported the

results of a close-to-realistic usability study with the

German “Defi Now!” app. Most of the participants

estimated that the app would be helpful in emergen-

cy cases. We recommended some improvements

based on best practices in various other first aid apps

and on our own observations during the usability

study.

First aid apps can increase their usefulness if

they offer easily accessible opportunities for learn-

ing and exploration. They should enable smartphone

owners to refresh their knowledge of first aid

measures, thus being better able to administer first

aid spontaneously when witnessing an emergency.

As for future work, this suggests the development of

first aid training apps that invite to dry-run first aid

procedures in a challenging manner.

REFERENCES

Cummins, R. O., Eisenberg, M. S., Hallstrom, A. P., &

Litwin, P. E. 1985. Survival of out-of-hospital cardiac

arrest with early initiation of cardiopulmonary resusci-

tation. The American Journal of Emergency Medicine,

3, 114-119.

Darley, J. M & Latané, J. M. 1968. Bystander Intervention

in Emergencies: Diffusion of Responsibility. Journal

of Personality and Social Psychology, 8, 377-383.

Gross, J. J. & Levenson, R. W. 1995. Emotion Elicitation

Using Films. In Cognition & Emotion, 9, 87-108.

Hewig, J., Hagemann, D., Seifert, J., Gollwitzer, M.,

Naumann, E., & Bartussek, D. 2005. A revised film

set for the induction of basic emotions. Cognition and

Emotion, 19, 1095-1109.

Jörg, G. 2012. Warum Menschen anderen nicht helfen.

See http://www.fh-heidelberg.de/?id=5236

Lange, T. 2011. Entwicklung eines Defibrillator-Verzeich-

nisses mit zugehöriger Smartphone-Applikation. Di-

ploma Thesis, University of Koblenz-Landau.

Bohm, K., Rosenqvist, M., Herlitz, J. Hollenberg, H. &

Svennson, L. 2007. Survival is similar after standard

treatment and chest compression only in out-of-

hospital bystander cardiopulmonary resuscitation.

Cicrulation, 116, 2908-2912.

Nielsen, J. 1994. Usability Engineering. Boston: Academ-

ic Press.

Otto, J. H. 2000. Methoden der Emotionsforschung:

Induktionsverfahren. In Otto, J.H.,

Emotionspsychologie, pp. 395-408. Weinheim: Beltz.

Rohrmann, B. 1978. Empirische Studien zur Entwicklung

von Antwortskalen für die sozialwissenschaftliche

Forschung. Sozialpsychologie, 9, 222-245.

Schleicher, R. 2009. Emotionen & Peripherpher-

physiologie. Lengerich: Pabst Science Publishers.

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