This study is the second in the series of studies on holistic approach
to analysis of medical data. Namely, besides the information regardin
g his/her disease, each hospitalized cancer patient also provides the
variety of data regarding his/her psychological, cultural, social, eco
nomical, genetic, constitutional and medical background. The aim of th
is series of studies is to introduce a model for holistic approach to
analysis of medical data, in this case clinical data regarding cervica
l cancer. The model requires the collection of as many such data as po
ssible for each patient in the sample, and after the satisfactory samp
le size is obtained (which should be at least five times greater than
the number of examined patient characteristics), the performance of fa
ctor analysis. In this particular study the authors have processed the
data regarding 25 characteristics of 1000 consecutive cervical cancer
patients treated between 1988 and 1990 at the Department for Gynecolo
gical Oncology of the University Hospital for Gynecology and Obstetric
s, Zagreb, Croatia. In factor analysis the principal components should
be rotated after the initial extraction (the authors recommend the us
e of oblimin rotation) in order to obtain better ground for interpreta
tion of the obtained results. The next step in the model is the stepwi
se exclusion of characteristics with smallest communalities according
to Kaiser-Meyer-Olkin criteria, and retaining the characteristics and
components with the most significant impact on the explained system va
riance. When the number of principal components and initial analyzed c
haracteristics is reduced to 3-4 and 7-10, respectively, the ultimate
interpretations and conclusions should be made.