Publication: Dependency evolution in Spanish disabled population : a functional data analysis approach
Loading...
Identifiers
Publication date
2013-02
Defense date
Advisors
Tutors
Journal Title
Journal ISSN
Volume Title
Publisher
Impact
Export
Abstract
In a health context dependency is defined as lack of autonomy in performing basic activities
of daily living that require the care of another person or significant help. However, this
contingency, if present, changes throughout the lifetime. In fact, empirical evidence shows
that, once this situation occurs, it is almost impossible to return to the previous state and in
most cases the intensity increases. In this article, the evolution of the intensity in this
situation is studied for the Spanish population affected by this contingency. Evolution in
dependency can be seen as sparsely observed functional data, where for each individual we
get a curve only observed at those points in which changes in the condition of his/her
dependency occur. We use functional data analysis techniques such as curve registration,
functional data depth or distance-based clustering to analyse this kind of data. This approach
proves to be useful in this context since it takes into account the dynamics of the
dependency process and provides more meaningful conclusions than simple pointwise or
multivariate analysis. The database analysed comes from the Survey about Disabilities,
Personal Autonomy and Dependency Situations, EDAD 2008, (Spanish National Institute of
Statistics, 2008). The evaluation of the dependency situation for each person is ruled in
Spain by the Royal Decree 504/2007 that passes the scale for assessment of the situation set
by Act 39/2006. In this article, the scale value for each individual included in EDAD 2008
has been calculated according to this legislation. Differences between sex, ages and first
appearance time have been considered and prediction of future evolution of dependency is
obtained
Description
Keywords
Chain-ladder, Dependency, Disability, Forecasting, Functional data, Time warping model