A Platform for Assessing Cancer Patients’ Quality of Life - New Contributions in Information Systems and Technologies - page 50

the response. The entire presentation layer is rendered on the client, reducing the need
for server processing.
The Quality of Life Monitor receives and records clinical data on the platform as
the patients respond, also on the platform, to the questionnaires. Two questionnaires
for evaluation of the quality of life were used: QLQ-C30 e H&N35. The Rasch model
was applied in order to have a score to assess the QoL. More details are described
next (sub-section C). It was also tested if there were differences between using the
platform and the support of paper in order to access the results of these questionnaires
and there were not statistical differences between the results using these two supports
[16]. The input data process has been simplified to typify the possible values for each
clinical patient variable and respective social demographic data. This procedure
facilitates the data entry process and reduces typing errors which simplifies the
posterior search for information.

QoL
Monitor

Module

Module

Module

Doctor

Server: weblogic
SGBD:Oracle

Patients

Fig. 1. Platform global architecture

The clinical practice of doctors and nurses allowed to determine a set of rules that,
depending on the answers given by the patient and the results obtained regarding their
QoL, enable the platform to provide guidance on the recommendation for the patient
(such as a specific therapy or psychological appointment). The inclusion of these rules
in the developed platform allows providing recommendations to the patients in short
term in order to improve their quality of life.

3.2 Rasch Model

Models of Item Response Theory (IRT) emerged to supply the needs resulting from
limitations of the Classical Theory of Measures (CTM) theory, which was
traditionally, and still is, used, in reviews. According to [17], IRT models, when
compared to CTM models, have several advantages, namely: the statistics are
independent of the items of the sample from which they were estimated; the results of
the examinees are independent of the difficulty of the test; the analysis of item
accommodates test items consistent with the level of knowledge of the examinees; the