ARREST 23: HartslagNu-app. A randomized study of two methods to alert volunteer rescuers

In 2005, the Arrest-study started to prospectively collect and analyse data of
out-of-hospital-cardiac-arrest victims, carried out by the department of
Cardiology of the Academisch Medisch Centrum.
In 2009, various dispatch center in the Netherlands started to use a text
message alert system (TM-alert system) for calls involving a suspected cardiac
arrest to alarm lay rescuers (TM-responders). Earlier research showed that in
12% of all out-of-hospital-cardiac-arrests (OHCA) an AED alerted via the
TM-alert system was connected to the patient. Median time from 112-alert to
defibrillation shock was 07:34, which was slightly faster than first-responders
(police i.e.) which had a median time to shock of 07:56. Also, the system
proved effective where needed: increasing survival for OHCA patients with a
shockable first rhythm in residential area.
The TM alert system uses predefined home- and/or work address and availability,
which does not guarantee that the person is, in reality, present at that
location at that time. This fact makes it uncertain how many of the selected
TM-responders are actually at the predefined location and being able to respond
to the TM-alarm call. To ensure there is an sufficient number of TM-responders
nearby, a large number of TM-responders (30 max.) is alarmed.
Some other countries are using alert systems which know the exact locations of
the lay rescuers from gps location of the smart phone of the lay rescuer. After
introducing a new smartphone app this will also be possible in the Dutch
setting. The smartphone app also grants other possibilities, such as showing a
map with the shortest route to the victim (possibly via an AED). Both changes
should enable a more effective and efficient response from the TM-alert system.

This study aims to determine if the alert of the lay rescuers is done more
efficient with the smart phone App compared to normal alert with a
text-message, avoiding alarming a redundant large group of lay rescuers and if
it shortens the time interval between emergency call and connection of an AED
from the TM-alert system at patients with an OHCA.

This study is a randomised controlled trial (RCT). When the dispatch center
receives a call with a suspected OHCA and activates the TM-alert system, the
TM-alert system automatically randomises the way of alerting. The TM-alert
system can alert TM-responders with a text-message based on predefined location
and availability (control group) or with a smartphone app based on actual
current location determined with geolocation of the smart phone (intervention
group). The study is conducted on intention-to-treat, which means that if there
are too few TM-responders with an activated smart phone app available, the
alerted group will be completed by TM-responders with text-messages, if
available. Data collection will prospectively be done through the existing data
collecting structure of the ARREST study.

For the study to succeed, as many TM-responders as possible should have the
smartphone app downloaded and activated, since the highest possible coverage is
desired. A minimum density of smart phone app users is necessary in order to be
able to see a real difference. For example, if only 10% of the volunteers would
activate the smartphone app the chance that we will see improvement is very
small. Therefore, the study will be preceded by an implementation phase. The
implementation phase will be used to determine the number of actually alerted
TM-responders using the smartphone app. The aim is to have at least 80% of the
alerted volunteers per alert in the intervention group to use the smartphone
app before the RCT starts.

The randomisation of the TM-alarmsystem will be active at the dispatch center
of North-Holland North and Twente. Each time when the TM alert system will be
activated, it will choose between one of the two ways of alarming:

Control group:
Alarming of TM-responders with text messages based on predefined location
(home/work) and availability.

Intervention:
Alarming of lay rescuers with smart phone-app based on actual location and
availability of the lay rescuers.

Alarmed TM-responders will receive a short questionnaire after each activation.
The treated patient will have no burden or risk associated with this randomized
alert study because their treatment does not differ between both randomized
groups.