Sofie Van Hoecke*, Electronics and Information Technology Lab, University College West-Flanders, Ghent University Association, Kortrijk, Belgium Wannes Kerckhove, Department of Information Technology, Ghent University, Gent, Belgium Thomas Dupont, Department of Information Technology, Ghent University, Gent, Belgium Filip De Turck, Department of Information Technology, Ghent University, Gent, Belgium Johan Decruyenaere, Department of Intensive Care, Ghent University Hospital, Gent, Belgium Track: Research Presentation Topic: Public (e-)health, population health technologies, surveillance Presentation Type: Oral presentation Submission Type: Single Presentation Building: MECC Room: 0.9 Athens Date: 2010-11-30 03:30 PM – 05:00 PM Last modified: 2010-09-21

If you are the presenter of this abstract (or if you cite this abstract in a talk or on a poster), please show the QR code in your slide or poster (QR code contains this URL).

Background
A computerized intensive care unit (ICU) is an extremely data-intensive environment, resulting in enormous databases. It is generally assumed that every patient generates around 16,000 different data values on a daily base. Not only the amount of data, but also the heterogeneity calls for automated data processing. Therefore, the Intensive Care Service Platfom (ICSP) has been developed which enables an automated surveillance of ICU patients. The patient’s data is processed by the ICSP medical support services and, depending on the priority, results are sent to the patient’s bedside terminal, the physician’s smart phone, or to an e-mail address.

Objective
An important factor to take into account while developing mobile applications is the battery lifetime. Up till now, ICSP only supported pull models for data retrieving on mobile devices. Consequently, the smart phones need to poll every few seconds for new data, resulting in a waste of resources so that the physicians need to recharge their mobile phone every few hours. Therefore this paper addresses the development of a mobile ICSP client application which combines a smart implementation of the pull model and push model, in order to prolong battery lifetime.

Methods
On one hand, a smart way of polling by using SMS is implemented. Whenever the ICSP has data to be delivered to a physician’s smart phone, an SMS is sent to that phone. This SMS is automatically processed, so that the smart phone’s Wi-Fi is set up and the new data is retrieved from the ICSP using the pull model by calling a REST web service. This way, Wi-Fi resources are only used while collecting patient data.
On the other hand, a push model is implemented where the smart phone passively waits until new patient data is pushed to it. In this case, the Wi-Fi needs to be on all the time and a web service needs to be running on the smart phone.

Results
Both scenarios have been implemented and tested for both the Google Android platform and Windows Mobile platform. The client application contains three tabs: one tab to manage subscriptions, another tab to actively query patient data, and one tab containing the received messages.
The advantage of the SMS pull model is that less resources are used, resulting in prolonged battery life. However, the cost of SMS is a disadvantage. Therefore, the final mobile client uses a hybrid version of both models in order to reduce this cost. The push model is used until battery lifetime drops below 30%. Whenever that happens, the client application switches to the SMS pull model and uses SMS to receive updates.

Conclusion
The mobile ICSP client application presented in this paper provides physicians with the timeliest patient data whenever and wherever needed so that physicians can focus on patient care. By combining push and SMS pull, the battery lifetime problem of current polling-based mobile solutions is solved.