![]()
31
| 2 | 2005
Klinisk Biokemi i Norden
tages ranged from 70 to 88%. The lowest percenta-
ges were attained during the weekends (Saturday
and Sunday 70%) and the highest percentages from
Mondays to Thursdays (86 - 88%), Fridays having
an intermediate percentage of 83%. This is in perfect
agreement with the variation in numbers of samples
per day. During the study period, there were about
17000 results per day from Mondays to Thursdays,
13733 on Fridays but only 1129 on Saturdays and
only 2734 on Sundays. There was more variation
in autoverification percentages between different
hours of the working day, ranging from 72 to 92%.
The highest values were attained in the middle of
the day and the lowest early in the morning and
late in the afternoon.
For autoverified results of the Hitachi Modular
analyzer, the 5th, 50th and 95th percentiles of the
verification TAT were 0, 0 and 1 minute, respecti-
vely. (The zero values and lack of decimals are due
to the fact that we can only measure the verifica-
tion TAT in minutes, not seconds.) For manually
verified results the 5th, 50th and 95th percentiles
of the verification TAT were 1, 8 and 30 minutes,
respectively.
When verification TATs were analyzed separately
for each day of the week, there was very little varia-
tion for autoverified results, the 5th and 50th per-
centiles being the same as for the whole study period
and the 95th percentiles ranging from 1 to 2 minutes.
For manually verified results, there was somewhat
more variation between days of the week, the 50th
percentiles ranging from 3 to 7 and the 50th percen-
tiles ranging from 20 to 22 minutes.
When verification TATs were analyzed separately
for times of the workday, there was again very little
variation for autoverified results, the 5th and 50th
percentiles being the same as for the whole study
period and the 95th percentiles ranging from 1 to
3 minutes. For manually verified results, there was
clear variation in the 50th percentiles that ranged
from 4 to 12 minutes and the 95th percentiles that
ranged from 16 to 37 minutes.
Discussion
According to our strategy the principal goal of our
laboratory is to produce reliable results in a timely
manner for fit-for-purpose laboratory tests keeping
processes simple.
This study shows that autoverification contributes to
these goals. Autoverification speeds up turnaround
times, especially by eliminating the particularly
long verification times that can occur from time to
time with manual verification. It clearly also makes
the process of reporting laboratory results simpler.
Although it was not something that could be shown
in this study, in our opinion autoverification also
contributes by making results more reliable. In
addition to simulating the way in which results are
manually verified, the autoverification program can
verify results by processes that would be tedious
Laboratory test
Number of
Percentage of
results
autoverified
results
Plasma creatinine
concentration
13645
86
Plasma C-reactive
protein
concentration
11424
89
Plasma potassium
concentration
10506
87
Plasma sodium
concentration
10190
88
Plasma alanine
aminotransferase
activity
8761
86
Plasma Aspartate
aminotransferase
activity
3939
84
Plasma alkaline
phosphatase
activity
4987
81
Plasma triglyceride
concentration
5172
86
Plasma cholesterol
concentration
4990
87
Plasma high-density
lipoprotein
cholesterol
concentration
4924
88
(Fortsætter side 32)
Table 3. Numbers of results and autoverification percent-
ages for the ten most common laboratory tests of the
selected Hitachi Modular analyzer during the four week
study period
