Safety Tips: Part Two – Redundancy By Gary Grist RN CCP Emeritus
John Nance And Redundant Systems
Many years ago, I heard a lecture by an aviation safety expert and accident analyst named John Nance. He said that the lives of the occupants of an airliner should not be dependent on a single mechanical system that, if it fails, could cost the lives of all those on board. He spoke about the redundant systems built into airliners, including having two pilots. Redundancy is the addition of critical back-up systems to increase reliability and safety.
No Redundant Systems On A HLM
As a perfusionist, I was struck by the realization that our patient’s (our passenger’s) life is dependent upon the mechanical systems and disposable components of the heart lung machine (HLM) – and that this machine is not designed with adequate redundant systems in case a primary system fails. The lone perfusionist is more likely to err or have difficulty troubleshooting without a copilot. In the first “Safety Tips” article, I spoke about two redundant systems for perfusionists: a back-up sweep gas system and a back-up arterial pump (1). The AmSECT Standard 6.1 does require a back-up gas supply during CPB without offering a specific description. In reality, how many perfusion programs have a redundant sweep gas system readily available and practice implementing it? The Standards and Guidelines make no mention of a back-up arterial pump for immediate use.
Example Of System Failure And No Redundant Backup
The failure to anticipate a problem and a lack of redundancy is illustrated by a recent Perfusion Improvement Reporting System II (PIRS II) report (2). A perfusionist working alone noticed that tubing within the arterial roller pump was splitting and leaking blood. He was not prepared to deal with this situation. He sent a nurse to the perfusion supply room to hunt for the materials he needed to fashion a new pump boot. Luckily the incident occurred near the normal termination of CPB, but still required the infusion of 1.5 L of crystalloid volume to make up for the blood loss from the split leak. After bypass, the perfusionist quickly fashioned a new pump boot and inserted it into the circuit should CPB need to be resumed. Fortunately, resumption was not necessary. Now realizing that an incident of this type is possible, the perfusionist is ordering preassembled sterile pump boots with the proper connectors to be kept in the program’s pump supply cart within arm’s reach of the HLM.
Learning From A Bad Experience
Learning from a bad experience like the incident described is not only unnecessary, it is dangerous to the patient. Certainly, a split pump boot is a well-known component failure for which every perfusionist should be prepared even though it is rare. Perfusionists should be able to anticipate most dangerous incidents BEFORE they occur by writing failure modes and effects analyses (FMEAs). These can tell what types of failures perfusionists can expect and how to prevent them or safely mitigate them. Many common CPB and ECMO FMEAs can be found on the AmSECT Safety Page along with information on how to write an FMEA. Secondly, perfusionists can learn from aviation safety analysts like John Nance and airliner manufacturers of the need for redundant systems so that failure of a single system will not jeopardize the lives of our “passengers”. I acknowledge that there are some incidents that FMEAs cannot anticipate like an earthquake (3). But perfusionists know about most of the things that can go wrong and can write FMEAs to prevent or mitigate them.
The Unsafe Perfusion Culture
I am not being critical of this perfusionist for being unprepared for this incident. He behaved as most perfusionists would in his situation. What I am criticizing is the culture within the perfusion profession that does not anticipate these types of failures and is not prepared to address them safely. Just because a program or individual perfusionist has never dealt with a dangerous incident does not mean they are practicing safely. (Most airliners never crash but they all have redundant systems, two pilots and escape hatches.) Safety practices need to be taken seriously and constantly improved whether they are ever used or not. The best way to demonstrate safety is by being prepared and practicing for many conceivable incidents BEFORE they happen.
One Answer To a Split Pump Boot
The perfusionist’s answer to the risk of a split pump boot is to have spare, sterile boots available on the supply cart within arm’s reach. That is one viable solution. Because this is such a rare incident, in all likelihood these units will expire before they are ever used again unless the perfusionists practice replacing the boot periodically. This requires a) a rigorous check system for spare equipment and b) periodic simulated pump boot replacement. However, there are alternatives which may be more practical.
An Alternative Answer
For example, the inflow tubing to the pump could be lengthened enough so that if the boot in the pump splits, the tubing could be advanced within the raceway, the defective section cut out and the advanced section connected in its place; in other words, redundant tubing. This is commonly done on ECMO roller pumps. But rather than waiting for the boot to split, the tubing is periodically “walked” to prevent the split from occurring. Even though the CPB pump does not operate within the long time frame that an ECMO pump does, the possibility of a split is very real, as this case suggests. Slack tubing can be built into the CPB circuit as a precaution. But nowadays many perfusionists cut all the slack out of their circuits to save prime volume, making it difficult or impossible to make a component change if needed. This may be advantageous at saving a little prime volume, but it is inherently unsafe.
A Better Alternative
Another alternative I prefer is to have a free-standing, battery operated centrifugal pump within arm’s reach. This pump can be designed to fit near the roller pump and if the tubing splits, the intact portions of the boot can be cut and reattached to the centrifugal pump head, priming it from the venous reservoir. This option actually addresses another potential problem: the centrifugal pump can be utilized if the primary arterial pump (roller or centrifugal) fails to operate.
Redundant Blood Flow Meter
Besides a redundant sweep gas system and a redundant arterial pump, three additional redundant systems should also be incorporated into the HLM and its management. The first is an independent blood flowmeter. Blood flow is arguably the most important measured parameter during CPB and should have a redundant system. Two recent PIRS II reports describe an unexpected drop in venous saturations. In both cases it was eventually determined that a slight kink in the line between the venous reservoir and the roller pump was responsible for a drop in the actual blood flow. As part of his troubleshooting efforts, one perfusionist gave an unnecessary blood transfusion before determining that a reduction in actual flow not detected by the roller pump flow readout was the real problem (4). The other perfusionist had a redundant blood flowmeter built into the system and was able to quickly determine the real cause of the fall in venous saturation and correct it quickly by repositioning the pump inflow tubing (5).
Redundant Oxygenator Change Out System
The “in series” change out of a failing or clotted oxygenator at a time when the patient is incapable of supporting his own circulation is unacceptable, particularly when a simple alternative is available.
The redundant system that is needed is a PRONTO line in every circuit (6,7). This is simply a shunt around the oxygenator which is normally kept clamped so it does not add to the dynamic prime. If the oxygenator should need to be changed out, a new unit can be placed “in parallel” using the PRONTO line without taking the patient off CPB. While not a complete redundant system in itself, the PRONTO line makes it possible to incorporate a redundant oxygenator without the risk of removing the patient from CPB.
Personnel Redundancy
The third redundant system should be another perfusionist or specially trained perfusion assistant who is immediately available in the theater with each case whenever it occurs (late night, weekends and holidays included). The AmSECT Guideline 15.1 (it should be a Standard) says: “The “n+1” staffing model should be utilized at all times, where “n” equals the number of operating/procedure rooms in use at any given time at a single site.” Only about 1/3rd of perfusion programs use the n+1 model on every case (8). In the case described above about the split boot, the case was on a weekend and there was no trained help who the perfusionist (and the patient) could rely on to quickly get the necessary supplies in an emergency. Instead, nurses who were unfamiliar with the needed supplies and where to find them were called upon for help. The perfusionist who was ill prepared to change the pump boot was just lucky the case terminated when it did. Or should I say the patient was lucky. Then again luck is not a safety plan. I understand the economic realities of additional personnel, but a single disastrous incident that could have been prevented or mitigated with better staffing might cost millions of dollars.
Revise Standard 18
AmSECT Standard 18 deals primarily with equipment maintenance, although it does call for the organization to have a plan for perfusion equipment failures and to have back-up supplies immediately available. But what types of plans are needed and what does “immediately available” mean? Individual programs can give this meaning by practicing and documenting the practice of their emergency plans often. I think that AmSECT Standard 18 should be revised to specifically describe the need for the five redundant systems I have described herein, as well as additional ones yet to be described. And those redundant systems should be included in the pre-CBP check list as being available and functioning properly.
Practicing Redundant Safety Techniques
Standard 18 calls for plans to deal with failures but makes no requirement to rehearse their implementation. A recent AmSECT poll showed that 31% of perfusion programs never practice or even discuss any emergency protocols. The implementation of each mechanical or component back-up should be practiced annually by each member of the perfusion team. A high fidelity simulation laboratory is not necessary to do this. For example, if there are five perfusionists on staff, then each perfusionist should practice replacing the roller pump boot in the operating room before discarding the used circuit after a completed case once each year with the other four perfusionists observing and critiquing. Similar exercises should be performed for the replacement sweep gas system, the implementation of a back-up arterial blood pump and the change out of an oxygenator using a PRONTO line. These practice sessions should be documented and available for review by internal as well as external authorities. Documentation should also show that a redundant blood flow meter is used and that qualified help (whose name is on the record) is immediately available. This is in line with the Joint Commissions Leadership Standard 5.2, which requires a proactive approach to patient safety.