Calibration guideline for the infusion pump analyzer applied in

secondary laboratories in Thailand

T. Chinarak1, K. Leetang1 and P. Wongthep1

National Institute of Metrology, Thailand (NIMT),

1

3/4-5Moo 3, Klong 5, Klong Luang, Pathumthani, 12120 Thailand

E-mail: [email protected]

Abstract

The flow laboratory of the National Institute of Metrology, Thailand (NIMT), in cooperation with the department of
medical sciences, and the department of health service support, ministry of public health developed the periodic testing
and maintenance procedures of the infusion pumps according to the international standards, IEC 60601-2-24 particular
requirements for the safety of infusion pumps and controllers, and requirements from the authorities in Thailand. The
testing procedures are now widely used in hospitals across country. Since the infusion analyzer is used as a tester in one
of the methods applied in the infusion pump testing procedures. Then, the calibration guideline for the infusion analyzer
has been developed to support the secondary laboratories for calibrating the devices. The guidelines provide
characteristic of the three main elements used for calibration which are the supply and flow control, the weighing system
and the data acquisition system. The infusion analyzer has to be calibrated at least three points; 50, 100 and 300 ml/h that
are required by the authorized person and normal conditions for patients in the hospitals. The uncertainties of calibrations
are range from 0.31% to 0.57%.

considerably sufficient for calibrating the infusion pump

1. Introduction analyzer which has the accuracy of ±1%. One reference
that emphasizes the uncertainty of this system is suitable
The infusion pump is one of the medical devices that is the MeDD report [2]. The report shows the
have been used widely in the hospitals. Generally, they uncertainty of the primary standard system of liquid
have the accuracy ranged from ±5% to ±10% depending flow facility at VSL is 0.11% (k=2).
on the IV set that they have been used with. In order to
confirm that the pumps are still working properly, the Noted that in this guideline, the pressure calibration
testing procedures are introduced. The Thai version of method is followed the Guideline DKD - R 6-1:
the testing procedures (ISBN: 978-616-11-2153-2) [1] Calibration of Pressure Gauges, however, the details of
are developed under the cooperation of the national this topic is not discussed in this paper.
institute of metrology (Thailand) (NIMT), the
department of medical sciences (DMSC), and the 2. Equipments
department of health service support (DHSS), ministry
of public health. The testing and maintenance The equipments used in the calibration system can be
procedures of the infusion pumps have been studied separated into three groups according to IEC 60601-2-
according to the international standards, IEC 60601-2- 24 [3] which are water supply, weighing system and
24 particular requirements for the safety of infusion data acquisition system. Each equipment is selected
pumps and controllers, and requirements from the based on the availability, user friendly and service from
authorities in Thailand. The testing procedures of the manufacturer.
infusion pumps have been distributed and applied to the
hospitals since 2014. 2.1Water supply
The water supply system is aimed for generating flow
In the testing procedures, the infusion pump analyzer is rates in the range of (10 – 1000) ml/h. The infusion
used for testing the infusion pump. That means that the pump is chosen to be the source of water supply. The
infusion pump analyzer is the important device. Then, infusion pump from Terumo, model TE171 using with
the next step of the project is the development of the the IV set for pump was selected to control the flow rate
infusion pump analyzer calibration procedures which for this system. The IV set is set to the pump and the
would fulfill the traceability. The calibration is aimed to inlet of the unit under calibration, UUC. At the outlet of
be as a guideline which has the least requirements for the UUC, the 1/4” stainless steel tube is connected and
the facility set by the secondary laboratories in the water is flowing through the tube to the weighing
Thailand. The calibration points are set at 50,100 and balance. The reverse osmosis, RO, water is used for this
300 ml/h but other points are optional. The uncertainty system.
of this setup is ranged from 0.31% to 0.57% that

FLOMEKO 2016, Sydney, Australia, September26-29, 2016 Page1

 1.Infusion Evaporation is another factor that has to be considered.
pump The evaporation rate is directly related to the mass loss
during the calibration period. Then, the evaporations are
UUC tested with four different conditions; largemouth of
polyethylene containers with and without cover, and
2. Weighing small mouth of polyethylene containers with and
System
without cover. Figure 3 reveals that the small mouth
with cover giving the best result among four conditions.
Moreover, the evaporation rate has to be counted into
the uncertainty due to mass measurement.
3. Data acquisition 2.3Data acquisition system
system
The data from the weighing balance are collected by
using the LABVIEW programme. The LABVIEW is
connected to the weighing balance (max. load 1020g)
Figure 1: The system setup via RS232 which is the real-time data processing. Both
measured mass and time are recorded every 30 seconds
2.2 Weighing system and in the form of text which can be analyzed by excel
The study of effects of weighing balances due to the or other suitable programme. The raw data can be
infusion devices [4] shows the factors that affect the calculated to find the discrete flow rate, the accumulate
uncertainty evaluations. Some of those factors are flow rate and trumpet curve. The successively sequence
recognized by the users who often use the balances. of data is collected for about one hour this is counted as
However, for the small flow rate calibration, more a one loop of calibration.
factors have to be considered such as evaporation rate
and environmental vibration. The Figure 2 shows the Moreover, other equipments used in this calibration are
shield that prevents the variation due to the laboratory temperature PT100 for measuring the water temperature
vibration and air flow. and the electronic thermo-hygrometer for measuring the
room conditions; room temperature and humidity. The
18G needle room conditions are controlled at (20 ± 2) ºC and (55 ±
Cover 15) %RH.
Evaporation
protection 3. Calibration Procedures

Shield The flow rates are calibrated at 50, 100 and 300 ml/h.
Variation of the
environment
Those flow rates are normally selected to infuse the
Fluid level
drugs or solutions to patients. At each flow rates, mass
and time are successively collected every 30 seconds
from the weighing system. The calibration period is one
Balance hour and repeated for three loops. The data from the
UUC are recorded depending on its software. However,
the UUC data have to show the measured flow rates.
Figure 2: Shield of weighing balance Other parameters such as the water temperature,
ambient temperature and ambient humidity are also
0.20 recorded.
0.18 0.17208
The measured mass and time are used for calculating
0.16 Largemouth w/o cover
the flow rates. The mass of each increment are
Evaporation rate (ml/h)

Largemouth w cover
0.14 converted to volume by using Equation (1) which comes
Smallmouth w/o cover
0.12 Smallmouth w cover from ISO/TR 20461 [5].
0.10
0.08 V 20=mw ×Z×Y
(1)
0.06
Where mw the collected mass of water
0.04 Z the combined factor for buoyancy correction
0.02088
0.02 0.00864
0.00436 and conversion from mass to volume
0.00 Y the thermal expansion correction factor of the
delivering device
Figure 3: Evaporation rate with four types of vessels
Since the measured mass is collected at every 30
seconds, then, the discrete flow rates are executed every

FLOMEKO 2016, Sydney, Australia, September26-29, 2016 Page2

30 seconds. Consequently, the discrete flow rates in one The Figure 4 to 6 shows the instability of the discrete
hour must be 120 data and the standard flow rates can flow rates from the standard system (weighing method)
be calculated from the average value of those data. and the UUC (infusion pump analyzer). The higher flow
rates, the higher instability. Hence, the instability of the
The flow rates of the UUC can only get from measured flow rates from UUC and the standard system
manufacturer’s software. The deviation of the set flow are included to the uncertainty evaluation.
rates is allowed at ±10%. In Figure 4 to 6, the discrete
flow rates of the UUC and standard system are plotted. 4. Uncertainty evaluation
The uncertainty evaluation for the small flow calibration
UUC flowrates
ml/h Standard flowrates can be written as shown below.
57 ml/h
57

uc=¿ √u2mw+u2Z+u2Y +u2t +u2instability_WB ¿+u2repeat +u2res +u2instability ¿¿

55 55
53 53
UUC UUC UUC (2)
51 51
49
The uncertainties of the mass to volume conversion
49 (umw, uZ, uY) can be derived by the methods as shown in
47
47
the ISO/TR20461 [1].
45 4.1 Time, ut
45
0 2000 The uncertainty due to time from the laptop, ut, is 0.010
0 1000 2000 3000
Time (s) Time (s) second. This value comes from the error occurred
between the accurate frequency generator and the
Figure 4: Discrete flow rates of UUC and standard system at 50 ml/h counting programmed in the laptop.

UUC flowrates Standard flowrates 4.2 Instability of the flowrate, ufluctuation_WB and
ml/h ml/h
110 110 uinstability_UUC,
The instability of the discrete flowrate of the standard
105 105 calibration system, ufluctuation_WB, can be considered from
the standard deviation of N data divided by the total
100 100 collected numbers of data.
SD Qstd
95 95 uinstability =
discrete

WB
√N (3)
90 90
The instability of the discrete flowrate of the UUC,
85 85 uinstability_UUC, can be considered from the standard
0 1000 2000 3000 0 1000 2000 3000 deviation of N data divided by the total collected
Time (s) Time (s) numbers of data.
SD QUUC
uinstability =
discrete
Figure 5: Discrete flow rates of UUC and standard system at 100 ml/h
UUC
√N (4)
UUC flowrates Standard Flowrates
ml/h ml/h 4.3 Repeatability, urepeat_UUC
295 295
urepeat_UUC, can be
290 290 The repeatability of the calibration,
considered from the maximum deviation of the error
285 285 caused by the calibration result.
error max−error min
280 280 urepeat =UUC
2 √3 (5)
275 275
4.4 Resolution, ures_UUC
270 270
ures_UUC, can be
0 1000 2000 3000 0 1000 2000 3000
Time (s) Time (s) The resolution of the calibration,
considered from the finest display of the unit under
Figure 6: Discrete flow rates of UUC and standard system at 300 ml/h calibration that user can read.

FLOMEKO 2016, Sydney, Australia, September26-29, 2016 Page3

 resolution result of trumpet curves is shown in Table 2 and Figure
uresUUC = 8. At flow rate 50 ml/h, the infusion pump has to infuse
2 √3 (6) the solution through the infusion pump analyzer for 31
minutes prior to start the testing. This number is applied
5. Results to channel 1 to 4 of the infusion pump analyzer.

The results of calibration of each set point is shown in Table 2 : The trumpet curve
Table 1. All channels of the infusion pump analyzer are Flow rate (ml/h)
Stability (window/min)
calibrated at flow rates; 50, 100 and 300 ml/h. The CH1 CH2 CH3 CH4
50 31 31 31 31
uncertainty of calibration are ranged from (0.31 – 100 11 11 11 11
0.57)% of reading. The error span of all channels (with 300 5 5 5 2
uncertainty) are within ±1% which meets the
specification of the analyzer. This information is considerably useful for the operator
in order to get the correct testing result when they are
Moreover, the calibration system was verified by using using the infusion pump analyzer to test the infusion
the standard flow meter (Bronkhorst, M13) which has pumps. Similarly, the stability time for 100 and 300
accuracy ±0.25%. The standard flow meter was set ml/h are 11 and 5 minutes that the infusion pumps have
instead of the UUC and tested at the same flow rates. to infuse the solution through the analyzer prior to start
The results of the calibration of the meter are compared the test.
with the result from its certificates and the international
comparison [2]. From those results of the standard flow 10.0

meter, the comparison result reveals that the En ratio is 5.0

less than 1 for all set flow rates which means that the
new calibration setup at NIMT has agreed performance Error (%) 0.0
0 5 10 15 20 25 30
as shown on Figure 7. -5.0

Table 1 : Calibration results of infusion pump analyzer -10.0

Window (mins)
Flow rate Channel Error Uncertainty 2.0
ml/h No. ml/h % (k =2)% 1.5
1.0
1 -0.08 -0.16 0.44
Error (%)

0.5
2 -0.03 -0.06 0.49 0.0
50 -0.5 0
3 -0.05 -0.10 0.47 5 10 15 20 25 30
-1.0
4 -0.20 -0.39 0.57 -1.5
1 0.16 0.17 0.31 -2.0
Window (mins)
2 0.28 0.29 0.33
100 1.5
3 0.37 0.39 0.47 1.0
4 0.14 0.15 0.42
Error (%)

0.5
1 0.47 0.17 0.31 0.0
2 1.52 0.54 0.31 -0.5 0 5 10 15 20 25 30
300 -1.0
3 1.47 0.51 0.33
-1.5
4 1.85 0.65 0.31 Window (mins)

0 50 100 150 200 250 300

1.5 Figure 8: Trumpet curve of flow rates; 50, 100 and 300 ml/h
M13WM
1
M13Com
6. Conclusion
Error (% )

0.5

0 The calibration guideline for the infusion pump analyzer

-0.5 has the least requirements for the standard system for
calibrating the infusion pump analyzer which has the
-1
accuracy of ±1% of reading. The calibration system –
Flowrates (ml/h)
-1.5 weighing based method – has uncertainty of the system
(0.31 – 0.57) % that is considerably adequate for
calibrating the infusion pump analyzer. The guideline
Figure 7: Uncertainty of measurement, error span and accuracy of
the standard flow meter (Bronkhorst, M13)
will be printed out and distributed to the secondary
laboratories in Thailand by the DHSS and DMSC.
Using the trumpet curve to consider the stability time
For future works, the proficiency test is required to
for testing the infusion pump of each flow rate is
confirm the equivalence of the ability of each
another important topic. The trumpet curve can be
laboratory.
calculated by followed the IEC60601-2-24 [3]. The

FLOMEKO 2016, Sydney, Australia, September26-29, 2016 Page4

References

[1] Standard procedures for testing of infusion pumps.

ISBN 978-616-11-2153-2.
[2] C. David et.al. “Project MeDD – Task 1.1
Comparison report” December, 2013
[3] IEC 60601-2-24: Medical electrical equipment –
Part 2-24: Particular requirements for the safety
of infusion pumps and controllers, 1998 – 02.
[4] D.M. Clarkson, “Accuracy estimations of testing
of infusion devices using weighing balances”,
Medical Engineering & Physics, 24, 229 – 235,
2002.
[5] ISO/TR 20461: Determination of uncertainty for
volume measurements made using the gravimetric
method, 2000 – 11- 01.

FLOMEKO 2016, Sydney, Australia, September26-29, 2016 Page5