445 / 492
ICCS S2-5
(P)
ACCURACY OF FLOW INDEXES BETWEEN DIFFERENT VOIDS IN THE SAME
PATIENT.
Israel FRANCO
1
, Jacob FRANCO
2
and Stephen YANG
3
1) New York Medical College, Section of Pediatric Urology, Chappaqua, USA - 2) STonybrook University Medical School,
Urology, Chappaqua, USA - 3) Buddhist Tzu Chi University School of Medicine, Urology, New Taipei, TAIWAN
PURPOSE
Attempts to normalize flows using nomograms or volume corrected flows have been made but with little success. We
recently derived a quadratic formula to estimate Q which is then used to generate a Flow index (FI) based on an
idealized normal voider. We set out to prove its reliability from one void to the next and amongst various bladder
volume and PVR scenarios.
MATERIAL AND METHODS
A data set consisting of 1268 children who underwent 2 voids was used to test our sex specific formulas to derive FI
Q
max
. and Q
avg
. Matched pairs of voids based on flow patterns, bladder volume, PVR were tested in various
permutations. Q
max
, Q
avg
FI for each void were compared and tested for accuracy using Root mean square error (√MSE)
and Median Absolute % error (medA%E). Non-parametric testing was done on the different groups to confirm no
difference from the first to the second void.
RESULTS
We found that the most accurate means of evaluating one flow from another was to use FI based off the Total bladder
volume and Idealized voider equation. Q
avg
and Q
max
were less accurate regardless of the type of permutations set up
with medA%E approaching 40% in disparate volume voids and dissimilar voiding patterns. Similar voiding patterns with
similar volumes and PVR were most accurate medA%E =14%, 15% and √MSE=0.23 and 0.27 females and males
respectively.
CONCLUSIONS
Even though non-parametric tests can show that there are no statistical differences between different ways to calculate
flow index and different groups of voiders. The best measure of how accurate a test is and how reproducible is to use
accuracy measures as we have done. We have shown that an idealized normal voider FI is reproducible from one flow
to the next and is a useful tool to follow children over time.