Findings
• Although environmental factors external to the pump were controlled and unchanged, a wide range of 8low variability was observed with some pumps • Two of the three large volume pumps administered 8luids at highly discontinuous rates, including periods of little or no infusion • Continuous infusion pumps exhibited ‘cyclic’ variations around the programmed 8low rate; Some pumps varied by more than +/-‐30% over time periods greater than 10 minutes • Only one large volume infusion pump was able to deliver 5% accuracy at the 2 minute observation window
Pump #3
0 20 40 60 80 100 120 Time (min)
Introduction
Ubiquitous throughout the hospital, large volume infusion IV pumps play a crucial role in delivering 8luids and complex doses of medications to patients in a wide range of care settings. Medications administered through the IV pump are assumed to infuse at a smooth and continuous 8low rate displayed on the screen. And, clinicians expect that the pump is delivering 8luid at the rate programmed. However, a number of factors in8luence IV pump 8low accuracy. Pump standards typically report 8low accuracy of +/-‐5% but this is in controlled laboratory conditions. In real world clinical conditions like those in hospitals today, many infusion pumps administer 8luids at highly discontinuous rates, including periods of little or no infusion. This study examined the short term 8low accuracy of three large volume infusion pumps and explores the clinical impact of 8low variability for high risk medications with short half-‐lives.
Conclusion
Medications administered at a speci8ied rate are assumed to infuse continuously and consistently at the programmed rate. However, the 8indings in this study prove this is not the case. • Although environmental conditions were controlled, two of the three studied pumps administered 8luids at highly discontinuous rates, reporting +/-‐30% variability over time periods greater than 10 minutes • The 8indings are especially meaningful if high risk medications with short half lives are delivered. Variations in 8low with short time periods could have a profound clinical impact on the patient • Discontinuous delivery of the medication may lead to a medical intervention including bolus, dose increase or decrease, and/or additional therapy • Results demonstrate that signi8icant differences exist in 8low continuity among infusion pumps and this variation could pose a risk of hemodynamic instability in an already compromised patient
Susan Niemeier MHA, BSN, RN Department of Clinical Excellence Ivenix, Inc., North Andover MA
Method
A 8low accuracy test was conducted with three large volume infusion pumps. The test strictly followed international industry standards.
• IEC 60601-‐2-‐24: Requirements for the safety of infusion pumps and controllers
• Infusion Pump Type 4: Continuous infusion 8low, non-‐continuous 8low
• Volumetric Infusion Pump: The delivery rate is set by the operator & indicated by the equipment in volume per unit of time
The study evaluated 8low accuracy of 0.5mL/hr .
Environmental conditions such as temperature, humidity, evaporation, vibration, and 8luid properties were 8ixed. Head height and back pressure were at the nominal conditions de8ined by the pump manufacturers.
Observed variability in the test results was due to the unique characteristics of the pumping mechanism for each large volume infusion pump.
A startup time plot (T0 – 8irst 60 minutes) was created to illustrate the instantaneous 8low rate versus time, for the 8irst two hours of the test. It’s intention is to show how long it takes the device to get to the desired 8low rate from zero and reveal 8low performance characteristics.
A trumpet curve was generated after the pumps stabilized (T1-‐second 60 minutes). The minimum and maximum % error in 8low rate is plotted versus the length of time used to average the data (2,5,11,19 and 31 min observation windows).
Questions? Comments? Please contact: [email protected]
Discussion
Medications with short half-‐lives need continuous administration to maintain therapeutic ef8icacy and desired clinical outcomes. These medications are mostly for the acutely ill patient, to either support or control blood pressure, control arrhythmias, provide sedation and in some cases to paralyze for therapeutic needs. For example, a septic patient who is extremely hypotensive might have norepinephrine administered to support blood pressure. Norepinephrine has a half-‐life of approximately 3 minutes with an even shorter duration of action of approximately 2 minutes. In essence, discontinuous delivery of the medication might lead to either the need to bolus (if appropriate), dose increase, and/or additional adjunct therapy. An infusion pump that continuously delivers the same amount of medication in a smooth and consistent fashion, versus small cycles, may lead to bene8icial patient outcomes.
Conversely, a patient with a hypertensive emergency requires drastic reduction in blood pressure. Medications such as nitroglycerin and sodium nitroprusside are widely used depending on the condition. Both medications have half-‐lives that
are less than two minutes as well as a duration of action that is about three to 8ive minutes. For instance, sodium nitroprusside has an extremely rapid action that quickly goes away with any interruption, such as reduced 8low or even an occlusion. Patients in a hypertensive emergency need careful but calculated reduction in blood pressure to avoid fatal outcomes such as stroke. As previously discussed, these patients cannot afford variations in medication delivery but rather accurate and consistent drug delivery.
AIMs
1) Compare the differences of 8low accuracy of three large volume pumps 2) Consider the therapeutic ef8iciency and clinical outcome of short half-‐life
medications if variations in the 8low rate occur
2 5 11 19 31 Time (min)
0 20 40 60 80 100 120 Time (min)
2 5 11 19 31 Time (min)
Flow
Rate %Error
0.5m
l/hr
0 20 40 60 80 100 120 Time (min)
A trumpet curve is a plot of min and max % error in 8low rate. Error in 8low rate (%) is the error in slope between the two points
Pump #3
Pump #1 Pump #2
2 5 11 19 31 Time (min)
2 minutes is the half life of nitroglycerin 2-‐3 minutes is the
half life of norepienephrine
Pump #1 Pump #2 +/-‐30% error >10 minutes
Wide swing over 10 minutes 5% accuracy achieved
The startup time plot is a 8low of instantaneous 8low rate. The curves illustrates T0 (start-‐up 0-‐60 min) and T1 (60-‐120 min) 8low vs time data with a 30 sec sample rate
Factors AffecJng Large Volume Infusion Pump Accuracy and Flow
References
Hurlbut, JC et al., In8luence of infusion pumps on the pharmacologic response to nitroprusside, Crit Care Med 1991 Jan:19(1):98-‐101
Klem, SA et al, In8luence of infusion pump operation and 8low rate on hemodynamic stability during epinephrine infusion, Crit Care Med 1993 Aug:21(8):1213-‐7
Krauskopf KH, et al., Disturbance of continuous, pump administration of cardiovascular drugs by hydrostatic pressure, Anaesthesist. 1996 May; 45(5): 449-‐52
Leff RD, Stull JC. Accuracy, continuity, and pattern of 8low from 8ive macrorate infusion pumps Am J Hosp Pharm. 1988 Feb;45(2):361-‐5
Vanderveen, T. "Averting highest-‐risk errors is 8irst priority." 2005. www.psqh.com/mayjun05/averting.html (19 Sept. 2005)
Weinger MB, Kline A. Re8lections on the Current State of Infusion Therapy. Biomed Instrum Technol. 2016;50(4):253-‐62
Acknowledgements
I’d like to thank Jesse Ambrosina, Ben Powers & Ivy Andreica for their invaluable guidance and support on this project.
Flow
Recommended
