PREVENTING PERFUSION INCIDENTS: A RISK REGISTER by Gary Grist RN CCP Emeritus
Playing It Safe
A recent Perfusion Improvement Reporting System (PIRS) report (8/12/2021) describes the changeout of an oxygenator because it appeared to have a manufacturing defect that was only noticed once blood entered the fiber bundle of the oxygenator (1). Although the oxygenator appeared to be functioning adequately, the decision was made to come off CPB because the oxygenator did not look “right.” Fortunately, this was before cross-clamping and while the patient was warm and capable of supporting himself. A long CPB time was expected and, even though the oxygenator was working adequately, the changeout was “playing it safe.”
Nothing New To Learn
The author writes: ”No preventing actions could have foreseen this problem under clear prime circumstances as this was potentially a structural issue with the oxygenator itself, only picked up once filled with blood.” This report is interesting because it details the commonsense reasoning and actions the perfusionist takes to deal with this problem. But is there anything new to be learned from this incident? Perfusionists everywhere did not have to wait for this report to tell them that an oxygenator can fail due to a manufacturing defect and other causes: clotting, high-pressure excursions, heat exchanger failure, cracked casing, sweep gas failure, and even a failure from an unknown cause.
Luck Is Not A Plan
One thing was learned about this particular perfusion program; they do not use a PRONTO line. Luck had it that the changeout could occur when the patient did not need CPB for support, in other words, a near miss. Luck, no matter how minor a role it plays in any situation, is not a plan. A less practiced perfusion team may have elected to wait to perform a changeout as speculated in the report: “This report exemplifies good team work insomuch as the decision making occurred before a much more time constrained emergency changeout may have been required.” By this time, all perfusion programs should be using a PRONTO line which enables an oxygenator to be changed out without taking the patient off CPB (2). Not having a PRONTO line is akin to not having escape hatches on jet airliners. With only a one-in eleven million chance of crashing, according to the FAA, chances are the escape hatches on any particular flight will never be used. But when the hatches are needed…they are needed! The risk of having a failed oxygenator that needs to be changed at a critical moment is orders of magnitude greater than an airliner crash. Not having a PRONTO line and not practicing using it is the height of irresponsibility of any perfusion program.
A Near Miss
The report described herein is an example of a near miss. But even a near miss is a failure at some level and often relies on luck to prevent a harmful incident. According to The New York State Near Miss Registry, “a near miss is an act of omission or commission that could have harmed the patient but did not as a result of chance (i.e., luck), prevention or mitigation” (3). The New York Near Miss Registry (NMR) is an online voluntary, confidential, anonymous reporting system for Internal Medicine interns and residents similar in purpose to the ANZCP PIRS.
An AmSECT PIRS?
I enjoy reading the PIRS reports, and I think Tim Willcox, as the editor of PIRS, does a good job of monitoring them (4). But the amount and quality of new information that can be gleaned for practical purposes are limited. Some have proposed that AmSECT design its own perfusion incident report system. There are two reasons I don’t think this should happen.
- American perfusionists are hesitant to report their mistakes no matter how deidentified the report may be. They are afraid of legal discovery, disapproval of their surgeons, anesthesiologists, and employers, and the chance that the origins of the incident will leak out to their fellow perfusionists. Witness these two examples. A Utah woman exsanguinated into a discarded CPB circuit without the perfusionist noticing. The incident was reported in the Washington Post newspaper (5). And the story of an inept perfusionist who was sued for millions because he could not find a hand crank when the arterial pump stopped at a critical moment. The incident is used for advertising the success of a malpractice law firm (6). Even Tim Willcox reported that the ANZCPs, who are much less vulnerable to civil litigation, are hesitant to report incidents due to a “defensive unit culture,” among other things.
- Another incident reporting system would be an unnecessary redundancy. Any English-speaking perfusionist wishing to report an incident can do so on the ANZCP PIRS. An AmSECT incident reporting system would be a copycat program and waste time and manpower that could be of benefit in developing a Risk Register.
An AmSECT Risk Register!
On the other hand, AmSECT could be the first to develop a national perfusion “Risk Register” as suggested in the Guide to Good Practice in Clinical Perfusion (7): “Clinical perfusion is a complex practice with recognized inherent risks. Local practices, procedures or circumstances which potentially increase these risks need to be identified, assessed and rated with mitigating action identified. Such risks need to be placed on the risk register of the provider organization and reported through the Clinical Governance, Risk and/or Audit Committees, depending on local arrangements. A risk workshop involving as many members of the team as possible, to ensure that all risks are understood and agreed across the team may be useful.”
The Risk Register and FMEAs
An AmSECT Risk Register could be populated with scores of FMEAs (Failure Modes and Effects Analyses) identifying the many perfusion risks, assessing them, describing preventative interventions or safe mitigation of the failures that do occur, and rating them as described in the Guide (8). FMEAs would provide more useful information over a broader range of incidents than Manufacturer and User Facility Device Experience (MAUDE) reports. Perfusion students could study the FMEAs as a means to efficiently pass down the knowledge and experience of dealing with perfusion risks to a younger generation. The register could be available for practicing perfusionists to consult and modify as needed for their individual local practices, procedures, or circumstances as suggested in the Guide. FMEAs could also be used for scenarios in simulations or tabletop discussions for risk management.
FMEAs, Peer Review And The Joint Commission
A group of experienced perfusionists would write and review FMEAs for perfusion risks and rank each risk according to a consensus. Proactive risk assessments like FMEAs are meant to prevent failures before they occur or mitigate them safely if they do occur and provide evidence that failures are being prevented. In addition, FMEAs satisfy the 2001 Joint Commission Leadership Standard LD 5.2: Support of Patient Safety and Medical/Health Care Error Reduction requirement for proactive risk assessments. A perfusion incident reporting system does not do this, nor does a patient safety organization (PSO).
Cost And Litigation
A PSO contains information which is de-identified and protected from legal discovery. However, the information is limited only to those members who pay an annual fee. Individual perfusionists would not be able to participate either because of the PSO rules or because of the cost. Even if AmSECT joined a PSO, only AmSECT members could participate. And eventually the cost to AmSECT members would be reflected in annual dues. An AmSECT Risk Registry would be free to all members because it would use the AmSECT web site as its platform and volunteer perfusionists as reviewers to minimize costs. I would encourage AmSECT to make the Risk Registry available to perfusionists world-wide as a community service. Could information in the registry FMEAs be used by plaintiff’s attorneys to support a case? Yes, that is a possibility in the same way that the AmSECT Standards and Guidelines could be used. But fear of litigation should not be a reason to limit generic perfusion safety information. Even without an FMEA example to support a case, expert perfusion witnesses could provide the similar testimony.
No Need To Wait For An Incident
New FMEAs could be added and peer-reviewed without waiting for an actual incident to occur. This is because FMEAs are designed to identify failures BEFORE they occur, including uncommon and common risks without implicating any actual patients or perfusionists. In fact, just the concept of a failure that has yet to occur could be described. For example, the quirkiness of a specific disposable product, an idiosyncrasy related to a particular brand of equipment, or a caution related to a new procedure could all be described in FMEAs before a dangerous incident actually occurs. A genuinely unique incident reported to the ANZCP PIRS would be handled in the same manner. The FMEA peer review process would determine by consensus the best way to prevent or deal with an emergency and to rate the risk from low to high. FMEAs in the register would be modified over time as procedures, disposables, and equipment changed.
A risk register can identify the weaknesses in any individual perfusion program. When deciding on resource allocation, it can convey to an administration the need for additional perfusion personnel, training, and equipment to reduce the risk. For example, redundant blood flow meters would reduce the risk of an inaccurate blood flow resulting from human error. A risk register could justify such a purchase. Or establish that a lone perfusionist is significantly risker than the presence of two trained people when dealing with critical failures such as arterial pump breakdown or oxygenator malfunction. One such failure can result in patient death and multi-million dollar litigation if not mitigated quickly and safely. A risk register could justify the expense of additional help.
Let’s Prevent Incidents, Not Report Them
Tim Willcox is doing a great job. Let him and the ANZCP PIRS continue to report incidents AFTER they occur and let an AmSECT Risk Register prevent incidents BEFORE they occur.