Primum Non Nocere: Patient Safety in Cardiopulmonary Bypass
Izumi Harukuni
Valerie Sera
TO ERR IS HUMAN
Perfusion safety is a problem that encompasses a broad range of topics for which there is no one solution. Since the first successful clinical use of cardiopulmonary bypass (CPB) by the pioneering cardiac surgeon John Gibbon in 1953, numerous positive changes in the engineering and manufacturing of equipment, surgical and anesthetic techniques, education of perfusionists, and the management of perfusion have made CPB safer. Separating patient safety during CPB and clinical outcome specifically related to CPB from the complex milieu of the perioperative course of cardiac surgery is a difficult task. For example, a study reported by Healey in 2002 identified an overall adverse event rate of 26.9% and a mortality rate of 3.34% among 1,438 cardiac surgical patients, including 1,596 individual procedures (1). As assessed by a rigorous peer review process, 49.5% of the 182 minor complications, 38.7% of the 181 major complications, and 25.0% of the 48 deaths in this study were categorized as avoidable. Although CPB was not specifically cited in this report, “equipment failure” was tabulated and had a zero occurrence rate for both complication categories and for mortality as well. Nevertheless, CPB-related adverse events do occur, are often avoidable, and can produce significant adverse events including fatalities.
Perfusion safety does not refer only to the incorporation of safety devices in the equipment and supplies used during CPB, rather it includes everything about safe conduct of CPB such as the heart-lung machines, circuit design, ancillary equipment, perfusion practices, surgical and anesthetic techniques, and perhaps most importantly, the training and education of perfusionists with emphasis on vigilance and communication with the other members of the team in the operating room. The literature contains a multitude of reports describing oxygenator failure (2,3,4,5), mechanical failure (6), electrical failure (7,8,9,10), massive air embolization (11,12), and preventative maintenance (13,14). There are accident surveys (15,16,17,18,19,20,21), comparisons of those surveys (22,23), reviews of the risks and hazards of CPB (24,25,26,27,28), review of safety devices (29,30), and other aspects of safe CPB (31,32). Defining the problem of perfusion safety requires identifying the types of incidents and their frequency through a quantitative approach (33).
Patient safety during CPB has been a prime concern since its earliest clinical history. In the earliest perfusion circuits and equipment, there were complex safety feedback loops and devices incorporated into the design. The first heart-lung machines had fail-safe mechanisms such as oxygenator blood level sensors, level floats, and pressure sensors, many of which were described in the classic perfusion textbook by Pierre Galletti in 1962, Heart-lung bypass; principles and techniques of extracorporeal circulation (34). Although the clinical roots of CPB are in the operative repair of congenital cardiac anomalies and valvular heart disease, arguably it was the rapid increase in the numbers of coronary artery bypass procedures, which followed the 1968 publication by Favaloro (35), that produced an explosive increase in the numbers of cardiac surgical procedures and the further development of CPB equipment, processes, and training programs. The need for the development of simple, reliable systems became a necessity. Disposable components for the bypass machines have largely replaced earlier reusable components. Oxygenator design has moved from large, clumsy film oxygenators through hard-shell bubble oxygenators and to the current disposable, single-use, self-contained membrane oxygenators most commonly used (in the USA) today. Improved electromagnetically coupled centrifugal pumps replaced the early roller pumps. Problems associated with air embolization and anticoagulation have been heavily investigated. More recently, reperfusion injury and the systemic inflammatory response elicited by the blood to foreign surface exposure inherent in current CPB have received substantial attention. Over the last six decades, there have been many improvements in the equipment and circuits used, but in the end, it is still difficult to mimic the anatomy and physiology of the heart and lungs. And in the attempt to do so, adverse events do occur. Examination of these occurrences with the intent of reducing their frequency is the focus of this chapter.
There is abundant literature regarding CPB with particular attention paid to CPB techniques and the equipment. The actual number of articles written specifically addressing perfusion safety is small in comparison, although an underlying theme of safety is inferred. Palanzo searched a proprietary
database (PerfSearch, property of American Academy of Cardiovascular Perfusion, and limited to use by its members) that specifically encompasses CPB-related manuscripts that appear in journals from cardiac surgery, anesthesiology, cardiology, perfusion, and biomedical engineering (33). Of more than 12,000 published manuscripts, the percentage that was dedicated to perfusion safety was only 1.4%, a total of 165 articles. Many of those 165 articles were case reports or small series that described air embolism. Others were reviews that included air accidents. Only 0.45% (54 of the total) reported on safety devices or issues different from gaseous emboli. Some of the earliest articles addressed the prevention of air embolization (36,37). The era of the late 1950s and 1960s, articles that discussed flow meters (38), safety devices for CPB (39), mechanical failure during CPB and a left ventricular vent valve (40) were published. The 1970s showed a decrease in the number of articles about perfusion safety, but popular topics during that time described aortic vents to prevent air embolization (41), retrograde dissection of the aorta during CPB (42), and run away pump head (43). The next decade witnessed an increase in the number of safety reports secondary to a large increase in the number of cardiac surgeons and training programs. Christman and Kurusz (29) published a review of safety devices available for preventing massive gas embolization. The safety devices were divided into four categories: devices attached to oxygenators, those devices that used proximal and distal to the pump head, and a device that functioned as a blood pump. There are reports of power failures (7,8) and reviews on the risks of CPB (25,26,27). Other topics reported were about preventive maintenance (13), oxygenator failure (2), safety devices for blood level regulation (44) and for pressure measurement (45,46). In 1988, the American Academy of Cardiovascular Perfusion devoted an entire issue in its journal Perfusion to the many aspects of perfusion safety. The 1990s focused on a different aspect of perfusion safety. Reports on risk management (32) and quality assurance (47) made their debut, and remain important aspects of CPB safety.
database (PerfSearch, property of American Academy of Cardiovascular Perfusion, and limited to use by its members) that specifically encompasses CPB-related manuscripts that appear in journals from cardiac surgery, anesthesiology, cardiology, perfusion, and biomedical engineering (33). Of more than 12,000 published manuscripts, the percentage that was dedicated to perfusion safety was only 1.4%, a total of 165 articles. Many of those 165 articles were case reports or small series that described air embolism. Others were reviews that included air accidents. Only 0.45% (54 of the total) reported on safety devices or issues different from gaseous emboli. Some of the earliest articles addressed the prevention of air embolization (36,37). The era of the late 1950s and 1960s, articles that discussed flow meters (38), safety devices for CPB (39), mechanical failure during CPB and a left ventricular vent valve (40) were published. The 1970s showed a decrease in the number of articles about perfusion safety, but popular topics during that time described aortic vents to prevent air embolization (41), retrograde dissection of the aorta during CPB (42), and run away pump head (43). The next decade witnessed an increase in the number of safety reports secondary to a large increase in the number of cardiac surgeons and training programs. Christman and Kurusz (29) published a review of safety devices available for preventing massive gas embolization. The safety devices were divided into four categories: devices attached to oxygenators, those devices that used proximal and distal to the pump head, and a device that functioned as a blood pump. There are reports of power failures (7,8) and reviews on the risks of CPB (25,26,27). Other topics reported were about preventive maintenance (13), oxygenator failure (2), safety devices for blood level regulation (44) and for pressure measurement (45,46). In 1988, the American Academy of Cardiovascular Perfusion devoted an entire issue in its journal Perfusion to the many aspects of perfusion safety. The 1990s focused on a different aspect of perfusion safety. Reports on risk management (32) and quality assurance (47) made their debut, and remain important aspects of CPB safety.
KNOWLEDGE FROM THE PAST
In the first 30 years of CPB, there was constant attention paid to promoting safety in perfusion as is illustrated by the previous accounting of reports describing all of the various safety devices that were engineered and manufactured. But, it was not until 1980 when the first hints appeared regarding the frequency of adverse incidents occurring during CPB. Table 23.1 is a summary of surveys documenting the types and frequency of incidents. The Stoney report was a survey of 349 cardiac surgeons between 1972 and 1977 in the United States and Canada to estimate the frequency of patients injured by accidental arterial line air embolism (15). It was the first comprehensive effort to determine the incidence of perfusion-related adverse events (AEs). The questionnaire asked several questions about accidents concerning the pump oxygenators, that is, episodes of arterial line air embolism, mechanical failure of the pump, electrical failure of the pump, disseminated intravascular coagulation (DIC) during or after bypass, and oxygenator failure. There were a total of 1,419 accidents that resulted in 100 permanent injuries and 264 deaths (Fig. 23.1).
Air embolism in the arterial line was responsible for 92 deaths, and DIC was responsible for 163 deaths. Mechanical, electrical, or oxygenator failures were responsible for nine deaths. Another common cause of accidents was the reversal of the tubing connected to the ventricular sump, causing air to be pumped into the left ventricle. When asked about the usage of low-level alarms, only 42% of the responders used them and only 20% used an alarm system that was equipped with an automatic pump shut-off. When asked about the use of an activated clotting time (ACT) during bypass only 37% did not, but interestingly, those using the ACT reported 313 episodes of DIC and those who did not use the ACT reported only 57 episodes of DIC. This is likely a nice example of the underreporting of adverse events that occur when surveillance for the occurrence is limited or absent. The authors concluded that most of the accidents caused by arterial line air embolism were a result of human error, but could be eliminated by consistent usage of low-level alarm systems and automatic shutoff devices. Additionally, the deaths and injuries secondary to DIC during CPB might be avoided by consistent use of ACTs or heparin titration. A possible disadvantage in the methodology of the survey was that the information gathered by the questionnaire relied on the memories of the cardiac surgeons and was not a review of past experience. It should be expected that this methodology could have produced an underestimation of the true incidence of accidents during CPB.
In 1981, Wheeldon published results from two surveys akin to the Stoney report (16