Infusion Pumps and Their Safety in the OR

Dr. Dave Schlotterbeck

Pump safety is central to ALARIS, the company formed from the merger of IVAC and IMED, two founders and consistent innovators in IV drug infusion, and I will offer an international perspective since our products in Europe are further advanced in the area of Total Intravenous Anesthesia. I will attempt to get at the fundamental questions that we had to answer, the biggest hurdles we had to master, and the tasks that are yet to be accomplished.

First, let’s look at the most significant changes over the past 15 – 20 years with respect to use of infusion pumps in the OR. The first infusion pumps began to show up in hospitals about 30 years ago when they were used primarily outside of the OR. Approximately half of the nutrition pumps were turned off during the cases because the anesthesiologists did not want to deal with them. Gravity infusion systems were the norm in the OR and infusion pumps typically appeared only if a patient was receiving IV nutrition. Infusion pumps began to show up primarily in heart cases and pediatric surgeries. Although anesthesia is a small market opportunity, anesthesia-specific pumps began to appear over 15 years ago. In addition, some of the general-purpose infusion pumps were marketed with features that were designed to make these devices more anesthesia-friendly. The recent growth in popularity of IV anesthesia has increased the use of infusion pumps in the OR, and this trend will continue to grow. In Europe, several syringe pump companies market TIVA (total IV Anesthesia) devices that incorporate drug recognition and pharmaco-kinetic dosing programs. The popularity of these systems has been enhanced through the efforts of the IV anesthetic drug manufacturers. Pumps are provided to the anesthesia departments to use in their proprietary drug system.

Next, we have to define what Anesthesiologists want in an infusion system. Anesthesiologists are the only physicians who routinely use infusion pumps because they desire ease of use and simple programming. In addition, certain features intended to make pumps safer to use outside the OR by staff with varying experience, training, and skill level were not desired by anesthesiologists. For example, anesthesiologists prefer to have an unlimited "Pause" or "Standby" mode with no advisory that the pause time has been exceeded. Nurses, on the other hand, want to be reminded that the infusion is paused and the device needs to be restarted. Anesthesiologists prefer large air detection limits, few, if any, programming confirmation steps, and "instant on" operation. Anesthesiologists also prefer thumb wheels, knobs, and dials to program infusions. Nurses would not consider such controls safe, since they could be operated easily by patients and their families. As far as disposable administration sets are concerned, anesthesiologists as a group are far less concerned about using needles, want considerably more access sites than nurses, and often prefer gravity infusions with standard roller rate adjustment clamps or non-electronic "dial-a-flow" type devices. Space considerations and lack of availability of power outlets have contributed to the popularity of multi-channel pumps.

Then, we must determine how the devices on the market today meet the needs of the anesthesiologist. For the devices used solely by the anesthesiologist, the consensus was they do meet the needs of the anesthesiologists. There are a small number of these devices, and they typically infuse directly from syringes. They are easy to set up, quick to operate, and programmed with the drugs and dosing (e.g., loading, bolus) used by the anesthesiologist. They do a good job of hitting the target. The disadvantage of these devices is a result of meeting the OR requirements – they are not intended to be used outside of the controlled OR environment where there is a 1:1, anesthesiologist/anesthetist per patient. The only exception is when the anesthesiologist is present in the ICU for a procedure. If one or more of the anesthesia syringe pumps are to be used, the pumps typically are carried to the ICU and returned after the procedure.

Unlike in Europe where syringe delivery is the norm for all but hydration and nutrition, syringes are infrequently used outside of neonatal and pediatric areas for continuous infusion of critical care drugs. Consequently, transfer of patients from the OR to either recovery or ICU requires the anesthesia syringe pump systems to be changed to nurse-friendly, general-purpose pumps which infuse from bottles and bags. If the drug infusion system has to be changed due to use of anesthesia pumps, there are significant costs, complexities and clinical considerations that can be avoided if the same devices, sets, and drugs can be used in both settings. For this reason, some of the general-purpose devices have incorporated an "Anesthesia" mode. In addition to setting performance parameters (e.g. air bubble size detected), this mode of operation allows the pump to be placed in an "Anesthesia Pause" or "Standby" mode with no time limits. When the device is unplugged from the AC receptacle, the pump automatically exits the anesthesia mode and reverts to a nursing "safe" mode. Approximately 70% of the anesthesia departments use general-purpose infusion pumps for cardiovascular drugs and other medications that will be used in the OR. This leaves a large number of hospitals still using two different systems. Many anesthesiologists continue to use the anesthesia syringe pumps for administration of anesthetic agents and muscle relaxants regardless of which devices are used.

Operating room use of general-purpose devices requires a compromise by anesthesiologists. Infusion device manufacturers continually add new capabilities as the marketplace redefines what is required including features that better meet the needs of the anesthesiologist. For example, a new feature on some devices allows the anesthesia department the option to create a special list of medications and dosing parameters that are set up for specific loading, bolus, and maintenance of drug infusions. The enhancements manufacturers address through software are much easier to address than those related to hardware, fundamental user interface/human factors, and drug container differences.

Finally, we need to identify the tasks that are yet to be accomplished. As an industry we will continue to wrestle with ways to meet the specific needs of the anesthesiologist with the same devices used by the nurses. In addition, our industry is attempting to deal with the problem of medication errors, especially those associated with programming and use of infusion devices. Improving safety of drug infusion, including use in the OR, will continue to be a priority for the infusion device manufacturers. It’s a balance between the "Keep it Simple" (KIS) principle and meeting the requirements brought about by increasingly complex drug therapy, fewer skilled nurses, and higher patient acuity. New technology such as bar code readers, personal digital assistants (PDA’s), and physician order entry by computer is surfacing in some hospitals, but we do not have a clear vision of how these advances will apply to the OR environment. In fact, this technology could make standardization between the OR and nursing units even more difficult, particularly when the systems used by nurses incorporate technology that is not useful or desired by anesthesia. Clearly this is an area where industry and APSF can collaborate to take a leadership role to insure the technology solutions are safe and functional.

User interface standardization is another area that needs to be addressed. Currently, there is no standardization of the user interface, even within companies. The competitive nature of the business, consolidation of pump companies, and the long useful life of our technologies all contribute to this issue. There is a compelling need to standardize terminology, programming, and capabilities to address cost, inventory management, training, and safety.

The growing acceptance of IV anesthesia and the availability of tested pharmacokinetic dosing models allow automating the drug infusion. As mentioned, there are several syringe pumps available in Europe that incorporate pharmacokinetic dosing algorithms with drug recognition. Also, there are syringe pumps marketed in Europe that provide the dosing algorithms for propofol and alfentanyl. FDA clearance for similar devices in the US could be an issue, although the open loop nature of the dosing plus the constant presence of the anesthesiologist should make this an easier task than our previous experience bringing a closed loop computer controlled system to market.

Outside of medicine, computerized closed loop control systems are widely used. Although a computer controlled sodium nitroprusside system was first described in the early 70’s, only one closed loop system has been marketed in the US. The IVAC Titrator¨ system was developed over 10 years ago to control nitroprusside in post-op heart cases. Although this system was very effective in controlling blood pressure, the system was only marketed for a short time. The developers have canceled more recent development projects including a pharmacological stress test and a heparin dosing system. The development hurdles are too high, and the investment too large to make these systems commercial successes in the current regulatory climate. However, open loop systems lower the regulatory and economic risks while providing many of the benefits of true closed loop systems.

For example, combining depth of anesthesia monitoring with anesthetic therapy via an open loop system is a possibility. Similar opportunities to combine drug infusion and patient monitoring exist in use of conscious sedation and pain control. We must work together to change the way the FDA looks at automation as we adopt systems that improve safety.

Infusion pumps have been marketed for over 20 years with computer ports and communication software, but we see little evidence of the use of this capability to monitor and document infusions. To date, this has been a useful feature for research in anesthesia, but it is not used in routine clinical practice. The information that is communicated from the pumps is limited to rates of infusion and volumes infused plus operating states such as alarms. As the sophistication of the drug programming increases, the amount and type of information available from the infusion pumps will increase. More study is required to determine the value of this information to the anesthesiologist and the most appropriate way to present and archive this information in computerized patient records.

A final area that needs further development is training. Our clinicians comment on the difficulty they have with inservice education for the anesthesiologists. Many anesthesiologists do not feel the need for inservice training, and others have 5 minutes or less to learn how to use new devices. The traditional methods used for nursing inservice are not effective for anesthesiologists, so we have to find a better way. We find that the lack of familiarity with the devices often results in features that were added for OR use not being used. In addition, they often experience frustration in dealing with alarms and frequently refuse to use devices that could continue with the patients as they are transferred out of the OR. Computer-based training may be more effective, but overall attitudes toward the need for training must change if we are to do a more effective job in training.

The infusion device market in anesthesia has developed over the past 30-plus years. Some of the devices are designed to be used only by anesthesia, while others are multipurpose devices intended to be used throughout the hospital. The multipurpose devices solve many cost and clinical issues, but often require a compromise on the part of anesthesiologists. Application of technology in hospitals is slow to be adopted, but holds the promise to resolve many of the issues. Computer monitoring and control, intelligent systems, and innovative training hold the promise of increasing efficiency while helping to reduce medication errors. Anesthesiologists are the only non-nursing group of clinicians using infusion devices routinely, and they use the devices in ways that are not typical of nursing use. Working together, the APSF and industry can set new standards that insure the highest level of safety while insuring both the anesthesiologists and nursing constituencies are considered.