| |
In a longitudinal
study subjects are followed over time typically with repeated
monitoring of symptoms or other variables. Such studies can vary
enormously in their size and complexity. At one extreme symptoms
could be studied repeatedly in a large group (or cohort) of patients,
from diagnosis to death. At the other extreme, symptoms among
a relatively small number of patients (a small cohort) could be
followed for a few days or weeks.
The table
below shows hospital stays for five patients, measured over a
period of 14 days. As before, the lines represent time spent in
the hospital. A longitudinal study takes place, recruiting patients
when they are admitted to hospital. Thus each patient is recruited
at point X, and followed for 7 days, irrespective of whether they
remain in the hospital or are discharged.
Figure
9.1: Longitudinal Design Example
|
|---|
 |
|
In
this example, symptoms are studied in all admissions. The first
symptom assessment describes the level and nature of symptoms
on admission. This type of data could not be obtained in a cross
sectional survey. The repeated assessment of symptoms will show
how symptoms change in the week after hospital admission.
Usually in longitudinal studies of symptoms a number of assessments
are planned over time. These can then be grouped to construct
summary measures. This approach reduces the set of n measurements
to a reduced number of measurements (often to a single value).
For example, in a study of 409 terminal patients, quality of life
was evaluated at referral, and then every week till death (Paci,
2001). Two summary measures were then calculated: the mean
of the QoL scores after admission and till the two weeks before
death, and the mean of the QoL scores in the two weeks before
death. The figure below shows an example of data from a longitudinal
study of symptoms and other problems, collected using the Palliative
Outcome Scale (Hearn
and Higginson, 1999).
Figure
9.2: Results of longitudinal assessment of symptoms
and problems in patients admitted to palliative care
programs
|
|---|
|
|
|