Aims
of Study:
To
assess recall
data reliability
and diary
data variability
on frequency
of voiding
or incontinence
episode.
Methods:
Symptomatically
stable patients
with urinary
frequency
and/or incontinence
kept a voiding
diary for
2 weeks. Before
and after
diary patients
recalled the
average numbers
of voiding
and incontinence
during the
precedent
2 weeks. The
collected
data were
analyzed for
the reliability
of recall
data and variability
of diary data
using SAS
system. The
average of
diary data
was regarded
as the gold
standard.
Applicable
mathematical
models and
parameters
were heuristically
explored to
fit the distribution.
To confirm
the applicability,
the diary
records of
32 urge incontinent
patients and
40 stress
incontinent
patients,
which had
been stored
previously,
were attempted.
Results:
The
cohort comprised
73 patients
(58 women
and 15 men)
aged 63.5
11.3 years
(meanSD),
of whom 49
complained
incontinence,
and the remaining
24, frequency
only. Recall
frequency
was larger
than gold
standard,
and the difference
was 0.52.1
(-6.3~4.7)
and 0.71.6
(-6.0~4.3)
(meanSD and
range) for
voiding and
incontinence,
respectively.
The difference
was consistent
before and
after diary.
Urinary frequency
appears to
follow Gauss
distribution
and was expressed
as 10.13.1
(mean SD). Mixed model indicated day-to-day variance as 1.98, giving 95%
confidence
interval for
individual
mean as [mean1.96*1.41/n
(here, n represents
the number
of observation
in days).
As for incontinence
frequency,
the variance
of individuals’
average was
estimated
as large as
11.1 with
a mean of
2.0. Given
individual’s
own average,
frequency
of incontinence
episode in
an individual
could be assumed
Poisson distribution
according
to the Poisson's
law of small
numbers. This
assumption
would be reasonable,
since individual’s
mean is equal
or more than
variance for
almost all
patients.
Using these
presumed parameters,
variability
(95% confidence
interval)
for frequencies
of voiding
and incontinence
were calculated
(Tables).
For example,
urinary frequency
of 8 calculated
by 3-day diary
recording
indicates
the true frequency
is 6.41to
9.59(81.59)
at a probability
of 95%. Similarly
an average
dairy incontinence
episode of
3 calculated
by 7-day observation
implies the
true value
between 1.72
and 4.28(31.28).
The confirmatory
study showed
the reasonable
fitting of
the model
and parameters
to both urge
and stress
incontinent
populations>Conclusions
Patients’
recall on
voiding frequency
and incontinence
episode would
not be sufficiently
reliable.
Urinary diary
enhances reliability,
and confidence
interval of
diary recordings
was given
in a relatively
simple fashion.
These intervals
should be
appreciated
in assessing
the symptom
severity and
comparing
a paired data
of specific
subjects.
Variability
of dairy voiding
frequency
|
Mean
|
Observation period (days)
|
|
|
3
|
7
|
10
|
14
|
|
Any value
|
1.59
|
1.04
|
0.87
|
0.28
|
Variability
of dairy incontinence
frequency
|
Mean
|
Observation period (days)
|
|
|
3
|
7
|
10
|
14
|
|
0.5
|
0.80
|
0.52
|
0.44
|
0.37
|
|
1
|
1.13
|
0.74
|
0.62
|
0.52
|
|
3
|
1.96
|
1.28
|
1.07
|
0.91
|
|
5
|
2.53
|
1.66
|
1.39
|
1.17
|
|
10
|
3.58
|
2.34
|
1.96
|
1.66
|
95% confidence
interval is
given as mean
table
figure.
