Original Article
A rapid infusion pump driven by micro electromagnetic linear actuation for pre-hospital intravenous fluid administration
Proc IMechE Part H: J Engineering in Medicine 2015, Vol. 229(2) 101–109 Ó IMechE 2015 Reprints and permissions: sagepub.co.uk/journalsPermissions.nav DOI: 10.1177/0954411914568692 pih.sagepub.com
Peng Zhao, Yinbao Chong, An Zhao, Lang Lang, Qing Wang and Jiuling Liu
Abstract A rapid infusion pump with a maximum flow rate of 6 L/h was designed experimentally using a micro electromagnetic linear actuator, and its effectiveness was evaluated by comparing with that of a commercial Power Infuser under preset flow rates of 0.2, 2, and 6 L/h. The flow rate, air detection sensitivity, occlusion response time, quantitative determination of hemolysis, and power consumption of the infusion devices were extensively investigated using statistical analysis methods (p \ 0.05). The experimental results revealed that the flow rate of the designed infusion pump was more stable and accurate, and the hemolysis was significantly less than that of the Power Infuser. The air detection sensitivity and the power consumption could be comparable to that of the Power Infuser except the occlusion response time. The favorable performance made the designed infusion pump a potential candidate for applications in pre-hospital fluid administration.
Keywords Micro electromagnetic actuator, infusion pump, flow rate, hypovolemia, fluid resuscitation
Date received: 17 June 2014; accepted: 11 December 2014
Introduction Hypovolemia, a common symptom in clinical manifestation and diagnosis, can lead to hypovolemic shock in burn and trauma cases.1 Patients with signs of hypovolemic shock caused by excessive loss of blood could be resuscitated initially in the ‘‘golden hours’’ if the intravenous (IV) fluids are transfused to the body tissues at the early determination of hypovolemia.2,3 Otherwise, some of the critically ill patients would die of multiple organ dysfunction syndrome (MODS) due to the decreased tissue perfusion.4 The transition to serious illness usually occurs during the critical ‘‘golden hours,’’ when definitive recognition and treatment in terms of outcome can provide maximal benefit. For treatment of hypovolemic shock, early and aggressive fluid resuscitation based on normalizing blood pressure is expected to improve perfusion of vital tissues and thereby survival.5 Typically, isotonic crystalloid is initially infused to elevate and restore intravascular volume as well as myocardial left ventricular end-diastolic volume toward a euvolemic state. Commonly, gravity or pressure infusion
is used in emergency departments for fluid resuscitation, but its low flow rate and the demand for more service members limit its applications especially in military and civilian settings.6 Alternatively, in-line mechanical infusion pumps can be used in the administration of blood products. They are typically driven by a step motor that moves several circularly arranged rollers. Due to the mechanical configuration of the actuator, this type of infusion devices is usually bulky and thus not portable for routine use in pre-hospital fluid administration. Rapid infusion devices with intelligent accessories can provide an infusion flow rate as high as 1.5 L/min.7 It is attracting much attention in pre-hospital fluid
Department of Medical Engineering, Xinqiao Hospital, Third Military Medical University, Chongqing, China Corresponding author: Yinbao Chong, Department of Medical Engineering, Xinqiao Hospital, Third Military Medical University, No.183, Xianqiao Central Street, Shapingba District, Chongqing 400037, China. Email: [email protected]
Downloaded from pih.sagepub.com at UNIV CALIFORNIA SANTA BARBARA on November 22, 2015
102
Proc IMechE Part H: J Engineering in Medicine 229(2)
administration for improving the outcomes for patients with signs of shock.8 Recently, rapid infusion devices have been widely investigated.9–11 However, these infusion devices have the following challenges in prehospital fluid resuscitation: (1) large volume, heavy weight, and cumbersome; (2) high consumption due to alternating current (AC) power supply; and (3) inconvenience with potential safety hazard. In addition, their viability and effectiveness were questioned in fluid administration for trauma patients.12 Comunale13 compared the performance between an infusion system FMS 2000 and a system Level 1 to check the feasibility. Therefore, the effectiveness, reliability, and stability of the rapid infusion devices deserve further research and improvement.14,15 The limitations of the rapid infusion devices for fluid resuscitation under high flow rates are closely related to the actuation motors. The Power Infuser as a new, portable, and rapid IV crystalloid infuser is approved and currently used for the treatment of hypovolemia.16 It is driven by a rotating step motor and rollers. This device has the advantages of small size, ease of use, low power consumption, and is capable of delivering fluid up to a rate of 6 L/h. In recent years, micro electromagnetic technology17 has been developed to realize the accurate transfusion of blood products in pre-hospital fluid administration. The micro electromagnetic linear actuator (MEMLA) is one type of fast linear oscillomotors that can convert different forms of energy into linear motion with low frictional loss.18 Due to the merits of portability, low power consumption,
miniaturization, and light weight, infusion pumps based on MEMLA would play a vital role in prehospital treatment of hypovolemia. But the literature reports on this issue are scarce. Therefore, it is of great significance to carry out a systematic investigation on this kind of rapid infusion device. In this work, a rapid infusion pump driven by MEMLA was first reported, with a maximum flow rate of 6 L/h. The features of infusion accuracy and infusion safety as well as power consumption of this infusion device were investigated comprehensively by comparing with those of the Power Infuser.
Methods Fluid resuscitation pump In this work, a rapid infusion pump driven by MEMLA was designed with five parts: pumps with built-in check valves, a MEMLA, an intelligent controller, air and occlusion detectors, and a power source module as shown in Figure 1. The inlet of the pump is connected to an infusion set, while the outlet is connected to a vein set through a standard catheter and injection needle. The pumps are fixed at each side of the stator of the MEMLA, which is driven by an N-S (North–South) alternating magnetic field. The pumps are squeezed by the stator operating in a certain frequency, and each pump is equipped with two check valves to control the fluid flow direction. The expected infusion flow rate is reached by adjusting the control voltage on the MEMLA through a closed-loop
Figure 1. Principle sketch of the rapid infusion device based on micro electromagnetic linear actuation.
Downloaded from pih.sagepub.com at UNIV CALIFORNIA SANTA BARBARA on November 22, 2015
Zhao et al.
103
Figure 2. Schematic diagram of a micro electromagnetic linear actuator: (a) stator and (b) top view of a multilayer planar coil designed using LIGA technology.
feedback algorithm for the preset rate and the actual flow rate. During the infusion process, air can be prevented from entering into the simulated body tissues by monitoring the impedance variation of the fluid at the outlet. By checking the mechanical load variation of the MEMLA, infusion will be stopped automatically when occlusion occurs at the inlet or outlet. In this case, the alarm will go off until the air and resulting occlusion are eliminated in order to maintain infusion safety. The power module was powered by a portable 9-V dry battery, and a low power consumption technique was adopted to extend the battery lifetime. The weight of this prototype device is 320 g, and some portion of its material will be replaced by lighter material. But its physical dimensions have not been completely determined. The designed pump can operate orderly by controlling the MEMLA. The driving force for fluid flowing originates from the extruding force of the MEMLA. Structuring of the MEMLA, including its suspension, was based on the transverse flux induction principle. The actuator is mainly composed of four parts: rotor, stator, outer stator, and the coils, as shown in Figure 2(a). The stator, in particular, is a multilayer planar coil consisting of a Si matrix, a NiFe alloy magnetic core, and a double layer coil, which is fabricated using LIGA technology to improve the coercive force of the permanent magnet. From the perspective of the physical, the stator terminal voltage can be expressed as a function of time (t) U(t) = I R(t) + L
di(t) +E dt
ð1Þ
where I represents the current through the stator windings; R and L signify the winding resistance and the winding self-inductance of the stator, respectively. The backward electromotive force (E) denotes the product of the electromagnetic force coefficient (ke) and the oscillating velocity (y) of the rotor E = Ke y
ð2Þ
According to the mechanical dynamics model, the extrusion force (F) on the infusion pumps can be expressed as follows F = B I L N = ma + my + kx + fa
ð3Þ
where m, a, and x are the mass, acceleration, and displacement of the rotor, respectively; m is the friction coefficient of the mechanical movement; k is a spring constant; fa is the air resistance; B is the magnetic flux density; L is the circumference of a stator winding; and N is the number of turns. Through controlling the movement velocity of the rotor, the infusion set can be extruded continuously, and finally, fluid infusion can be performed. In addition, the Power Infuser commonly used in the current pre-hospital IV infusion administration was taken as a reference infusion pump in this work. It can be operated under six tunable settings from 0.2 to 6 L/h through a switch. This device is a battery-powered mechanical pump with a sterile and disposable cartridge through which the fluid flows. The core of the mechanical pump is a direct current (DC) micro-motor, which is powered by a 12-V lithium battery or an alternative 12-V AC/DC adapter. The cartridge consists of a short segment of standard IV tubing, connectors, and an air-permeable membrane that filters the unwanted air within the system. It also has a built-in alarm to indicate the occurrence of occlusion.
Experimental details In this work, two prototype pumps were selected randomly and labeled as Series A to test the flow rate, air and occlusion detection, hemolysis, and power consumption. Another two Power Infusers labeled as Series B were provided as control infusion pumps. In the test, the preset flow rate was set at 0.2 L/h (minimum), 2 L/h (medium), or 6 L/h (maximum). Each test was repeated 10 times, except for the free hemoglobin (fHb) determination in hemolysis that was repeated 5 times.
Downloaded from pih.sagepub.com at UNIV CALIFORNIA SANTA BARBARA on November 22, 2015
104
Proc IMechE Part H: J Engineering in Medicine 229(2)
Figure 3. Schematic description of the experimental system for the infusion performance test. The sealed plastic container with air chambers was used to simulate a 70-kg-weight adult with a blood volume of approximately 6 L.
To design a safe and available infusion pump, some key performance parameters of the device should be qualified. The fHb concentration in packed red blood cells (PRBCs) should be not more than 200 mg/dL. The response time for the occlusion alarm should be less than 30 s. The mean flow rate error should be less than 15%.
Measurement of flow rate and power consumption. The experimental system is presented schematically in Figure 3. The infusion container, infusion pump, and test chamber were placed on the same plane. Ringer’s lactate solution (RLs) placed in the infusion container was used as the blood substitution for transfusion. The infusion container was connected to the infusion pump using infusion set, and the outlet was connected to the test chamber using a standard Luer connection and an 18 G injection needle. The operation voltage and current of the infusion pump was monitored by a programable DC power IT6322 and a PV6300 suite, and then the power consumption can be calculated. The sealed test chamber was designed with a volume of 6 L to simulate the IV infusion tissues,19 and the central venous pressure (CVP) of the chamber was set at a fixed value of 0.49 kPa. The test chamber was placed on an electronic platform scale of Model CPA 16001S for weight monitoring. Real-time data of the chamber weight can be recorded every 30 s by our self-developed data acquisition software based on Microsoft Visual C++ Studio 2010. By monitoring the weight variation of the plastic chamber, the infusion flow rate can be computed using the following formula y t = 60 1=r dW(t)=dt
ð4Þ
where yt (L/h) is the infusion flow rate, r (g/L) and W (g) are the density and the weight of RLs, respectively, and t (h) is time. Determination of infusion safety. Infusion safety measurements involve air detection, occlusion monitoring, and quantitative determination of hemolysis. For air detection, the test was performed as follows: turn on the infusion pump, preset a desired flow rate, and fast inject air with bubble sizes of 50, 40, 30, 20, or 10 mL successively at a location 10 mm away from the impedance detector, so that the air can be detected. Successful detection results in the air being prevented from entering into the simulated body tissues, causing the system to stop running when the control panel indicates an air alarm. Occlusion detection was carried out similarly to the air detection, but using a Reference Pressure Monitor (RPM4) to monitor the occlusion and record the alarm response time and the maximum occlusion pressure. The concentration of fHb in PRBCs was measured using Trinder reaction for quantitative analysis of hemolysis. For hemolysis evaluation, a standardized relationship between the concentration of the fHb solution and its optical absorbance should be derived first. In this, the fHb standard solutions were prepared using a dilute concentration ranging from 0.1 to 2.0 g/L in even steps of 0.1 g/L. Subsequently, the absorbance (A) of each fHb solution was measured at 505 nm wavelength using a spectrophotometer. According to the Lambert–Beer law, A can be written as follows Io A = ln ð5Þ It
Downloaded from pih.sagepub.com at UNIV CALIFORNIA SANTA BARBARA on November 22, 2015
Zhao et al.
105
Figure 4. (a) Infusion flow rate versus time under a preset rate of 0.2 L/h for Series A and B, (b) infusion flow rate versus time under a preset rate of 2 L/h for Series A and B and (c) infusion flow rate versus time under a preset rate of 6 L/h for Series A and B.
where Io and It are the incident light intensity and the transmitted light intensity, respectively. The functional relation between the concentration (y) of the standard fHb solution and its absorbance A is roughly fitted using the least-squares method with R2 of 0.995 y = 212:94A + 1:36198
ð6Þ
In the test, each sample of 5 mL taken respectively from the PRBCs before and after transfusion was prepared to extract the supernatant fluid with fHb content by separation of red cells using a Centrifuge. By measuring the absorbance of the supernatant fluid, we can obtain the concentration of the fHb solution in accordance with the above equations (5) and (6). Finally, the hemolysis of this system can be evaluated. The PRBCs used in the test were supplied by Xinqiao Hospital, Third Military Medical University, with a storage period less than 35 days.
Statistical method The experimental data in this work were analyzed based on SPSS V13.0 software. Data about infusion flow rate
and power consumption were studied using variance analysis. The difference in the occlusion response time, the maximum occlusion pressure, mean power consumption, and hemolysis between this rapid infusion device and the Power infuser was analyzed using a ttest, while the air detection sensitivity was analyzed using the x2-test. Here, a significant difference is defined as p \ 0.05.
Results and discussion Accuracy of infusion flow rate Time-dependent flow-rate curves for Series A and B were compared under the different preset flow rates, as shown in Figure 4(a)–(c). The measurements were carried out in a period T1 of 2 h at even intervals of 10 min. The average flow rates for both infusion systems are summarized in Table 1. As shown in Figure 4(a) and (b), the average flow rate of Series A just fluctuates around the preset value with a small deviation, while for Series B, it takes on a rather large deviation from the preset value, even though the final flow rate tends toward it. Under high preset flow rates of 6 L/h, the
Downloaded from pih.sagepub.com at UNIV CALIFORNIA SANTA BARBARA on November 22, 2015
106
Proc IMechE Part H: J Engineering in Medicine 229(2)
Table 1. Mean and standard deviation of flow rates for Series A and B under different preset flow rates. Preset flow rate
0.2 L/h
2 L/h
6 L/h
Series A Series B
0.20 6 0.04 0.24 6 0.03
2.00 6 0.22 2.22 6 0.23
5.42 6 0.77 4.96 6 0.60
Statistical significance of p-values under different preset flow rates is 0.001.
Figure 5. (a) Percentage variation of Ep (max) and Ep (min) for the designed infusion system against observation window duration over the analysis period of T2. The overall percentage error of EA is also presented. The preset rate is set at 2 L/h. (b) Percentage variation of Ep (max) and Ep (min) for the Power Infuser against observation window duration over the analysis period of T2. The overall percentage error of EB is also presented. The preset rate is set at 2 L/h.
average flow rates for both infusion systems are lower than the preset value. This is because that the infusion device cannot respond instantly to the fast changing internal pressure of the test chamber under a high preset rate, consequently restraining the flow rate. The mismatch between the infusion cartridge and the injection needle size can be another reason. Thirdly, the dropping mechanical energy transformation efficiency in the case of high preset flow rate causes that the infusion flow rate can not reach the preset value. Through calculation, it was found that the mean flow rate error for the designed pump was 0%, 0%, and 9.7%, corresponding to the three preset flow rates, while for the Power Infuser, the mean flow rate error was 20%, 11%, and 17.3%, corresponding to the three preset flow rates. Generally, the flow rate error not more than 15% is permitted for a safe infusion pump. Therefore, the designed infusion system is rather accurate than that of the Power infuser system, and the flow rate of 2 L/h could be a suitable value for IV transfusion. In addition, according to our variance analysis, the two sets of data for Series A and B show a statistically significant difference with a p-value of 0.001. Trumpet curves were also employed to evaluate the accuracy of the infusion flow rates for the rapid infusion system, as shown in Figure 5. The basic safety of the infusion pumps is assessed according to the standard of IEC 60601-2-24:2012,20 which is the particular
requirement for the basic safety and essential performance of infusion pumps and controllers. The first 31 min of the second hour of T1 are chosen as the period T2, and data recorded within this period for a preset rate of 2 L/h were extracted to analyze the maximum error Ep (max) and the minimum error Ep (min). The sampling frequency is 30 s. The observation window p is set at 1, 2, 5, 11, 19, and 31 min 8 9 j + ps1 X m
A rapid infusion pump driven by micro electromagnetic linear actuation for pre-hospital intravenous fluid administration
Proc IMechE Part H: J Engineering in Medicine 2015, Vol. 229(2) 101–109 Ó IMechE 2015 Reprints and permissions: sagepub.co.uk/journalsPermissions.nav DOI: 10.1177/0954411914568692 pih.sagepub.com
Peng Zhao, Yinbao Chong, An Zhao, Lang Lang, Qing Wang and Jiuling Liu
Abstract A rapid infusion pump with a maximum flow rate of 6 L/h was designed experimentally using a micro electromagnetic linear actuator, and its effectiveness was evaluated by comparing with that of a commercial Power Infuser under preset flow rates of 0.2, 2, and 6 L/h. The flow rate, air detection sensitivity, occlusion response time, quantitative determination of hemolysis, and power consumption of the infusion devices were extensively investigated using statistical analysis methods (p \ 0.05). The experimental results revealed that the flow rate of the designed infusion pump was more stable and accurate, and the hemolysis was significantly less than that of the Power Infuser. The air detection sensitivity and the power consumption could be comparable to that of the Power Infuser except the occlusion response time. The favorable performance made the designed infusion pump a potential candidate for applications in pre-hospital fluid administration.
Keywords Micro electromagnetic actuator, infusion pump, flow rate, hypovolemia, fluid resuscitation
Date received: 17 June 2014; accepted: 11 December 2014
Introduction Hypovolemia, a common symptom in clinical manifestation and diagnosis, can lead to hypovolemic shock in burn and trauma cases.1 Patients with signs of hypovolemic shock caused by excessive loss of blood could be resuscitated initially in the ‘‘golden hours’’ if the intravenous (IV) fluids are transfused to the body tissues at the early determination of hypovolemia.2,3 Otherwise, some of the critically ill patients would die of multiple organ dysfunction syndrome (MODS) due to the decreased tissue perfusion.4 The transition to serious illness usually occurs during the critical ‘‘golden hours,’’ when definitive recognition and treatment in terms of outcome can provide maximal benefit. For treatment of hypovolemic shock, early and aggressive fluid resuscitation based on normalizing blood pressure is expected to improve perfusion of vital tissues and thereby survival.5 Typically, isotonic crystalloid is initially infused to elevate and restore intravascular volume as well as myocardial left ventricular end-diastolic volume toward a euvolemic state. Commonly, gravity or pressure infusion
is used in emergency departments for fluid resuscitation, but its low flow rate and the demand for more service members limit its applications especially in military and civilian settings.6 Alternatively, in-line mechanical infusion pumps can be used in the administration of blood products. They are typically driven by a step motor that moves several circularly arranged rollers. Due to the mechanical configuration of the actuator, this type of infusion devices is usually bulky and thus not portable for routine use in pre-hospital fluid administration. Rapid infusion devices with intelligent accessories can provide an infusion flow rate as high as 1.5 L/min.7 It is attracting much attention in pre-hospital fluid
Department of Medical Engineering, Xinqiao Hospital, Third Military Medical University, Chongqing, China Corresponding author: Yinbao Chong, Department of Medical Engineering, Xinqiao Hospital, Third Military Medical University, No.183, Xianqiao Central Street, Shapingba District, Chongqing 400037, China. Email: [email protected]
Downloaded from pih.sagepub.com at UNIV CALIFORNIA SANTA BARBARA on November 22, 2015
102
Proc IMechE Part H: J Engineering in Medicine 229(2)
administration for improving the outcomes for patients with signs of shock.8 Recently, rapid infusion devices have been widely investigated.9–11 However, these infusion devices have the following challenges in prehospital fluid resuscitation: (1) large volume, heavy weight, and cumbersome; (2) high consumption due to alternating current (AC) power supply; and (3) inconvenience with potential safety hazard. In addition, their viability and effectiveness were questioned in fluid administration for trauma patients.12 Comunale13 compared the performance between an infusion system FMS 2000 and a system Level 1 to check the feasibility. Therefore, the effectiveness, reliability, and stability of the rapid infusion devices deserve further research and improvement.14,15 The limitations of the rapid infusion devices for fluid resuscitation under high flow rates are closely related to the actuation motors. The Power Infuser as a new, portable, and rapid IV crystalloid infuser is approved and currently used for the treatment of hypovolemia.16 It is driven by a rotating step motor and rollers. This device has the advantages of small size, ease of use, low power consumption, and is capable of delivering fluid up to a rate of 6 L/h. In recent years, micro electromagnetic technology17 has been developed to realize the accurate transfusion of blood products in pre-hospital fluid administration. The micro electromagnetic linear actuator (MEMLA) is one type of fast linear oscillomotors that can convert different forms of energy into linear motion with low frictional loss.18 Due to the merits of portability, low power consumption,
miniaturization, and light weight, infusion pumps based on MEMLA would play a vital role in prehospital treatment of hypovolemia. But the literature reports on this issue are scarce. Therefore, it is of great significance to carry out a systematic investigation on this kind of rapid infusion device. In this work, a rapid infusion pump driven by MEMLA was first reported, with a maximum flow rate of 6 L/h. The features of infusion accuracy and infusion safety as well as power consumption of this infusion device were investigated comprehensively by comparing with those of the Power Infuser.
Methods Fluid resuscitation pump In this work, a rapid infusion pump driven by MEMLA was designed with five parts: pumps with built-in check valves, a MEMLA, an intelligent controller, air and occlusion detectors, and a power source module as shown in Figure 1. The inlet of the pump is connected to an infusion set, while the outlet is connected to a vein set through a standard catheter and injection needle. The pumps are fixed at each side of the stator of the MEMLA, which is driven by an N-S (North–South) alternating magnetic field. The pumps are squeezed by the stator operating in a certain frequency, and each pump is equipped with two check valves to control the fluid flow direction. The expected infusion flow rate is reached by adjusting the control voltage on the MEMLA through a closed-loop
Figure 1. Principle sketch of the rapid infusion device based on micro electromagnetic linear actuation.
Downloaded from pih.sagepub.com at UNIV CALIFORNIA SANTA BARBARA on November 22, 2015
Zhao et al.
103
Figure 2. Schematic diagram of a micro electromagnetic linear actuator: (a) stator and (b) top view of a multilayer planar coil designed using LIGA technology.
feedback algorithm for the preset rate and the actual flow rate. During the infusion process, air can be prevented from entering into the simulated body tissues by monitoring the impedance variation of the fluid at the outlet. By checking the mechanical load variation of the MEMLA, infusion will be stopped automatically when occlusion occurs at the inlet or outlet. In this case, the alarm will go off until the air and resulting occlusion are eliminated in order to maintain infusion safety. The power module was powered by a portable 9-V dry battery, and a low power consumption technique was adopted to extend the battery lifetime. The weight of this prototype device is 320 g, and some portion of its material will be replaced by lighter material. But its physical dimensions have not been completely determined. The designed pump can operate orderly by controlling the MEMLA. The driving force for fluid flowing originates from the extruding force of the MEMLA. Structuring of the MEMLA, including its suspension, was based on the transverse flux induction principle. The actuator is mainly composed of four parts: rotor, stator, outer stator, and the coils, as shown in Figure 2(a). The stator, in particular, is a multilayer planar coil consisting of a Si matrix, a NiFe alloy magnetic core, and a double layer coil, which is fabricated using LIGA technology to improve the coercive force of the permanent magnet. From the perspective of the physical, the stator terminal voltage can be expressed as a function of time (t) U(t) = I R(t) + L
di(t) +E dt
ð1Þ
where I represents the current through the stator windings; R and L signify the winding resistance and the winding self-inductance of the stator, respectively. The backward electromotive force (E) denotes the product of the electromagnetic force coefficient (ke) and the oscillating velocity (y) of the rotor E = Ke y
ð2Þ
According to the mechanical dynamics model, the extrusion force (F) on the infusion pumps can be expressed as follows F = B I L N = ma + my + kx + fa
ð3Þ
where m, a, and x are the mass, acceleration, and displacement of the rotor, respectively; m is the friction coefficient of the mechanical movement; k is a spring constant; fa is the air resistance; B is the magnetic flux density; L is the circumference of a stator winding; and N is the number of turns. Through controlling the movement velocity of the rotor, the infusion set can be extruded continuously, and finally, fluid infusion can be performed. In addition, the Power Infuser commonly used in the current pre-hospital IV infusion administration was taken as a reference infusion pump in this work. It can be operated under six tunable settings from 0.2 to 6 L/h through a switch. This device is a battery-powered mechanical pump with a sterile and disposable cartridge through which the fluid flows. The core of the mechanical pump is a direct current (DC) micro-motor, which is powered by a 12-V lithium battery or an alternative 12-V AC/DC adapter. The cartridge consists of a short segment of standard IV tubing, connectors, and an air-permeable membrane that filters the unwanted air within the system. It also has a built-in alarm to indicate the occurrence of occlusion.
Experimental details In this work, two prototype pumps were selected randomly and labeled as Series A to test the flow rate, air and occlusion detection, hemolysis, and power consumption. Another two Power Infusers labeled as Series B were provided as control infusion pumps. In the test, the preset flow rate was set at 0.2 L/h (minimum), 2 L/h (medium), or 6 L/h (maximum). Each test was repeated 10 times, except for the free hemoglobin (fHb) determination in hemolysis that was repeated 5 times.
Downloaded from pih.sagepub.com at UNIV CALIFORNIA SANTA BARBARA on November 22, 2015
104
Proc IMechE Part H: J Engineering in Medicine 229(2)
Figure 3. Schematic description of the experimental system for the infusion performance test. The sealed plastic container with air chambers was used to simulate a 70-kg-weight adult with a blood volume of approximately 6 L.
To design a safe and available infusion pump, some key performance parameters of the device should be qualified. The fHb concentration in packed red blood cells (PRBCs) should be not more than 200 mg/dL. The response time for the occlusion alarm should be less than 30 s. The mean flow rate error should be less than 15%.
Measurement of flow rate and power consumption. The experimental system is presented schematically in Figure 3. The infusion container, infusion pump, and test chamber were placed on the same plane. Ringer’s lactate solution (RLs) placed in the infusion container was used as the blood substitution for transfusion. The infusion container was connected to the infusion pump using infusion set, and the outlet was connected to the test chamber using a standard Luer connection and an 18 G injection needle. The operation voltage and current of the infusion pump was monitored by a programable DC power IT6322 and a PV6300 suite, and then the power consumption can be calculated. The sealed test chamber was designed with a volume of 6 L to simulate the IV infusion tissues,19 and the central venous pressure (CVP) of the chamber was set at a fixed value of 0.49 kPa. The test chamber was placed on an electronic platform scale of Model CPA 16001S for weight monitoring. Real-time data of the chamber weight can be recorded every 30 s by our self-developed data acquisition software based on Microsoft Visual C++ Studio 2010. By monitoring the weight variation of the plastic chamber, the infusion flow rate can be computed using the following formula y t = 60 1=r dW(t)=dt
ð4Þ
where yt (L/h) is the infusion flow rate, r (g/L) and W (g) are the density and the weight of RLs, respectively, and t (h) is time. Determination of infusion safety. Infusion safety measurements involve air detection, occlusion monitoring, and quantitative determination of hemolysis. For air detection, the test was performed as follows: turn on the infusion pump, preset a desired flow rate, and fast inject air with bubble sizes of 50, 40, 30, 20, or 10 mL successively at a location 10 mm away from the impedance detector, so that the air can be detected. Successful detection results in the air being prevented from entering into the simulated body tissues, causing the system to stop running when the control panel indicates an air alarm. Occlusion detection was carried out similarly to the air detection, but using a Reference Pressure Monitor (RPM4) to monitor the occlusion and record the alarm response time and the maximum occlusion pressure. The concentration of fHb in PRBCs was measured using Trinder reaction for quantitative analysis of hemolysis. For hemolysis evaluation, a standardized relationship between the concentration of the fHb solution and its optical absorbance should be derived first. In this, the fHb standard solutions were prepared using a dilute concentration ranging from 0.1 to 2.0 g/L in even steps of 0.1 g/L. Subsequently, the absorbance (A) of each fHb solution was measured at 505 nm wavelength using a spectrophotometer. According to the Lambert–Beer law, A can be written as follows Io A = ln ð5Þ It
Downloaded from pih.sagepub.com at UNIV CALIFORNIA SANTA BARBARA on November 22, 2015
Zhao et al.
105
Figure 4. (a) Infusion flow rate versus time under a preset rate of 0.2 L/h for Series A and B, (b) infusion flow rate versus time under a preset rate of 2 L/h for Series A and B and (c) infusion flow rate versus time under a preset rate of 6 L/h for Series A and B.
where Io and It are the incident light intensity and the transmitted light intensity, respectively. The functional relation between the concentration (y) of the standard fHb solution and its absorbance A is roughly fitted using the least-squares method with R2 of 0.995 y = 212:94A + 1:36198
ð6Þ
In the test, each sample of 5 mL taken respectively from the PRBCs before and after transfusion was prepared to extract the supernatant fluid with fHb content by separation of red cells using a Centrifuge. By measuring the absorbance of the supernatant fluid, we can obtain the concentration of the fHb solution in accordance with the above equations (5) and (6). Finally, the hemolysis of this system can be evaluated. The PRBCs used in the test were supplied by Xinqiao Hospital, Third Military Medical University, with a storage period less than 35 days.
Statistical method The experimental data in this work were analyzed based on SPSS V13.0 software. Data about infusion flow rate
and power consumption were studied using variance analysis. The difference in the occlusion response time, the maximum occlusion pressure, mean power consumption, and hemolysis between this rapid infusion device and the Power infuser was analyzed using a ttest, while the air detection sensitivity was analyzed using the x2-test. Here, a significant difference is defined as p \ 0.05.
Results and discussion Accuracy of infusion flow rate Time-dependent flow-rate curves for Series A and B were compared under the different preset flow rates, as shown in Figure 4(a)–(c). The measurements were carried out in a period T1 of 2 h at even intervals of 10 min. The average flow rates for both infusion systems are summarized in Table 1. As shown in Figure 4(a) and (b), the average flow rate of Series A just fluctuates around the preset value with a small deviation, while for Series B, it takes on a rather large deviation from the preset value, even though the final flow rate tends toward it. Under high preset flow rates of 6 L/h, the
Downloaded from pih.sagepub.com at UNIV CALIFORNIA SANTA BARBARA on November 22, 2015
106
Proc IMechE Part H: J Engineering in Medicine 229(2)
Table 1. Mean and standard deviation of flow rates for Series A and B under different preset flow rates. Preset flow rate
0.2 L/h
2 L/h
6 L/h
Series A Series B
0.20 6 0.04 0.24 6 0.03
2.00 6 0.22 2.22 6 0.23
5.42 6 0.77 4.96 6 0.60
Statistical significance of p-values under different preset flow rates is 0.001.
Figure 5. (a) Percentage variation of Ep (max) and Ep (min) for the designed infusion system against observation window duration over the analysis period of T2. The overall percentage error of EA is also presented. The preset rate is set at 2 L/h. (b) Percentage variation of Ep (max) and Ep (min) for the Power Infuser against observation window duration over the analysis period of T2. The overall percentage error of EB is also presented. The preset rate is set at 2 L/h.
average flow rates for both infusion systems are lower than the preset value. This is because that the infusion device cannot respond instantly to the fast changing internal pressure of the test chamber under a high preset rate, consequently restraining the flow rate. The mismatch between the infusion cartridge and the injection needle size can be another reason. Thirdly, the dropping mechanical energy transformation efficiency in the case of high preset flow rate causes that the infusion flow rate can not reach the preset value. Through calculation, it was found that the mean flow rate error for the designed pump was 0%, 0%, and 9.7%, corresponding to the three preset flow rates, while for the Power Infuser, the mean flow rate error was 20%, 11%, and 17.3%, corresponding to the three preset flow rates. Generally, the flow rate error not more than 15% is permitted for a safe infusion pump. Therefore, the designed infusion system is rather accurate than that of the Power infuser system, and the flow rate of 2 L/h could be a suitable value for IV transfusion. In addition, according to our variance analysis, the two sets of data for Series A and B show a statistically significant difference with a p-value of 0.001. Trumpet curves were also employed to evaluate the accuracy of the infusion flow rates for the rapid infusion system, as shown in Figure 5. The basic safety of the infusion pumps is assessed according to the standard of IEC 60601-2-24:2012,20 which is the particular
requirement for the basic safety and essential performance of infusion pumps and controllers. The first 31 min of the second hour of T1 are chosen as the period T2, and data recorded within this period for a preset rate of 2 L/h were extracted to analyze the maximum error Ep (max) and the minimum error Ep (min). The sampling frequency is 30 s. The observation window p is set at 1, 2, 5, 11, 19, and 31 min 8 9 j + ps1 X m
