| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Eur J Cardiothorac Surg 2006;29:645-652
© 2006 Elsevier Science NL
EACTS Invited Guest Lecture |
a Department of Thoracic and Cardiovascular Surgery, Cleveland Clinic, Cleveland, OH 44195, USA
b Department of Quantitative Health Sciences, Cleveland Clinic, Cleveland, OH 44195, USA
Received 29 November 2005; received in revised form 3 February 2006; accepted 7 February 2006.
* Corresponding author. Address: Department of Thoracic and Cardiovascular Surgery, Mail Code JJ-40, Cleveland Clinic, Cleveland, OH 44195, USA. Tel.: +1 216 444 6712; fax: +1 216 445 0444. (Email: blackse{at}ccf.org).
Abstract
For each risk factor for Black Death he discovered, John Graunt in 1662 made a recommendation for action, such as fleeing the night air brought to the City of London docks on foreign ships. These stopped the plague for 200 years! In this honored guest lecture, I focus on thinking beyond the risk factors to their neutralization by effective action without waiting for perfect knowledge of mechanisms. For example, the decreasing risk of repairing atrioventricular septal defect resulted from risk factors that stimulated focused research to devise a better operation. General reduction in risk reduced mortality of aortic valve replacement in patients with chronic aortic regurgitation and severe left ventricular dysfunction. Optimal timing of surgery after stabilizing patients reduced risk of the Norwood procedure. Managing intractable cardiogenic shock reduced risk-adjusted mortality in post-infarct ventricular septal defect. However, we must think beyond traditional risk factors to neutralization of those brought by patient characteristics, imponderables, treatment delivery systems, institutions delivering care to groups of patients, and society caring for its population. There are limitations to thinking beyond risk factors. First, neutralizing strategies require improving processes or introducing new ones. Yet there may be process–outcome disconnect, or even counterintuitive or contradictory relations among outcomes and what are thought to be best processes. Nevertheless, thinking beyond risk factors to their neutralization by research, innovation, and application of knowledge can be as spectacularly successful as was fleeing foul air in halting Black Death.
Key Words: Risk factors • Outcomes research • Clinical epidemiology
1. Introduction
In 1603, as the Black Death was rampaging across Europe, the City of London began publishing each Thursday a record of christenings and burials [1]. However, those who read these records ‘made little use of them than to look at the foot, how the burials increased or decreased, and among the casualties, what has happened rare, and extraordinary, in the week current [2,3].’
Unlike those who stopped at counting anecdotes, shopkeeper John Graunt believed the Bills of Mortality could be used to reveal the nature of Black Death and suggest ways to stop it. For each risk factor identified, he recommended an action (Table 1 ). City officials listened, and Black Death disappeared 200 years before these risk factors were understood.
|
The point I want to emphasize is that Graunt thought beyond the risk factors. He translated risk factors into effective action without waiting for perfect knowledge. Today, I want to explore the value of thinking beyond the risk factors for failure of surgery to achieve its desired results.
2. Thinking beyond the risk factors to their neutralization
Framingham investigators coined the term ‘risk factors [4,5].’ By the mid-1960s, they realized that the cause of heart disease was as elusive as that of Black Death 300 years earlier. Instead, they focused on weak associations with heart disease – hypertension, hyperlipidemia, smoking – calling them ‘risk factors.’ They hypothesized that heart disease might be reduced by population-wide risk factor modification. A coronary risk calculator was distributed across America in the hope that heart disease could be neutralized, just like the plague, by educating the public in primary and secondary prevention [6].
In 1979, Dr John Kirklin introduced our ‘incremental risk factor concept’ in his presidential address before the American Association for Thoracic Surgery [7]. In its simplest form, an incremental risk factor is a variable that reflects increased difficulty in achieving surgical success [8,9]. It is not a cause–effect mechanism! The adjective ‘incremental’ meant that the association of a risk factor with surgical failure was positive, adding an increment of risk over and above that of all other factors [10].
My focus today is on neutralizing incremental risk factors in cardiothoracic surgery. Is it rational to believe that they can be neutralized to prevent or reduce surgical failures, even if mechanisms of their association with outcome are incompletely understood?
Before answering that question, one must ask, ‘What is surgical failure?’ We define it as not achieving the desired planned or attempted goal of an operation [11]. Some claim that any failure to achieve a desired end should be considered an error [12]; this, in my judgment, is harsh and unhelpful. Instead, we recognize two primary causes of surgical failure: lack of scientific progress to prevent it, and failure to apply knowledge, which we do consider error (some have termed this preventable vs unpreventable adverse outcome [13]). By separating these causes according to the present state of knowledge, we can propose two programmatic strategies for neutralizing risk factors: (1) generate new knowledge by research and innovation, and (2) apply available knowledge to minimize occurrences and consequences of inevitable human error. Let me illustrate with four examples.
2.1 Neutralizing risk factors for death after AV septal defect repair
Babies with AV septal defect have a number of immutable risk factors, most prominently their genetics (trisomy 21, chromosome deletions, non-Down syndromes, and others). However, at one time, young age at operation was a risk factor. Is age immutable, or is it a surrogate for other, potentially neutralizable factors? As shown in Studer et al. [14], young age was neutralized between 1973 and 1980 (Fig. 1
). Some of you vividly remember the challenge in the 1970s of operating on babies! Lucio Paranzan organized the ‘Under Three Months’ meeting in Bergamo, and Aldo Casteneda and Brian Barratt-Boyes, among others, championed neonatal surgery.
|
Neutralizing age as a risk factor required overcoming many technical barriers, among which was incomplete repair or failed repair of the AV valve [14]. How could this risk factor be neutralized? There were many surgical possibilities, but the disease was too uncommon to conduct randomized trials. We needed new knowledge and innovative thinking. So Nevin Katz went to the lab. Using a tensinometer, he demonstrated substantial differences in tensile strength of various suturing techniques [15]. Reinforced sutures were far stronger than non-reinforced ones. Thus, an important risk factor was able to be directly neutralized by research and ‘bench-to-bedside’ implementation of a better technique.
2.2 Neutralizing risk factors for death after valve replacement for aortic regurgitation and severe left ventricular (LV) dysfunction
Another way to neutralize risk is indirectly, by overall risk reduction measures. Such was the case of neutralizing risk factors for death after surgery for aortic regurgitation combined with chronic, severe LV dysfunction [16]. This is a condition for which some advocate cardiac transplant.
Fig. 2 shows hospital mortality for aortic valve replacement with or without severe LV dysfunction over a 30-year period. By 1985, it was indistinguishable for patients with low or high ejection fraction (EF). In fact, we statistically demonstrated that low EF was still a risk factor with unabated strength, but its influence had been indirectly neutralized by general programmatic decrease of risk of valve replacement.
|
|
Thus, until a disease mechanism is understood, we may be able to effectively neutralize some risk factors by consistent use of available general knowledge.
2.3 Neutralizing early risks in aortic atresia
A third risk-neutralizing strategy is optimal surgical timing, which is illustrated by management of newborns with aortic atresia by Congenital Heart Surgeons Society institutions [17]. Greatest risk of death is early after a Norwood-type operation. At a number of institutions, these babies were operated on emergently to save the child's life. Yet when we looked at the cardiac transplant experience for aortic atresia, we saw that the majority of these babies could be medically managed with prostaglandins for a protracted period (Fig. 4
). Neutralizing risk factors did not require abandoning anatomic repair, but rather, optimizing the patient's preoperative condition to permit elective intervention on a stable individual. This strategy of optimal timing of operation claims a few lives, because about 5% of babies cannot be stabilized, but it salvages far more.
|
Hospital mortality after repair of post-infarct VSD has not changed from about 30% at Cleveland Clinic in 30 years (Lam BK, Gillinov AM, Rajeswaran J, Blackstone EH. Personal communication). But that is not the whole story. Over that 30 years, time from myocardial infarction to surgery has dropped from months to days (Fig. 5A). Patients months out from developing a VSD have strong tissue, not mush; surgery is easy and secure. Further, nowadays essentially all patients are in cardiogenic shock (Fig. 5B). So, if mortality has remained constant, dramatic improvements in managing desperately ill patients must have occurred. Indeed, if we take a constant case mix predicted to have 80% mortality in 1969, these cases would have 10% morality in 2000. Thus, was the surgeon too quick to pronounce our patient unsalvageable based on his past memory and not on today's results?
|
Perhaps the surgeon was remembering patients like one from the mid-1980s who was in shock, with aneuria and right ventricular dysfunction. At that time, we would have predicted near zero 30-day survival. There would have been good reason to conclude he was unsalvageable. However, if the same patient were seen today, his predicted 30-day survival 20 years later would be 78%. Although risk is still high, the situation is not hopeless.
So, what can we say about neutralizing the risk of cardiogenic shock? It requires
No doubt alternative management strategies are on the horizon. But these days, because of dramatic risk factor neutralization, few patients with post-infarct VSD are unsalvageable.
3. Thinking beyond the usual risk factors
So far, I have focused on only one outcome—death. Complications, failure to rescue from complications, and adverse impact on quality of life are others that need to be reduced by research and innovation, and by developing effective, practical processes based on knowledge.
Just as we need a wide view of surgical failure, we also need a wide view of risk factors. These include patient characteristics at operation, imponderables (chance and lack of knowledge), treatment delivered to individual patients, the institution delivering care to groups of patients, and society delivering health care to its population.
3.1 Patient risk factors
Patient risk factors at operation include chronic health status, acute physiologic status, and the cardiothoracic disease on which we are focused, including its character, extent, severity, acuity, and accelerating and compensating factors. Does multiplicity of these factors ever make a common operation such as primary isolated coronary artery bypass grafting (CABG) inadvisable?
We investigated this question in a subset of 27,000 patients operated on from 1971 to 1998. Twenty patient risk factors were identified: age >75 years, female gender, left main disease, three-system disease, previous myocardial infarction, LV dysfunction, mitral regurgitation, tricuspid regurgitation, heart failure, ventricular arrhythmia, atrial fibrillation, smoking, chronic lung disease, diabetes, peripheral vascular disease, prior stroke, carotid disease, hypertension, renal failure, and hyperthyroidism. Over 4000 patients had four or fewer of these comorbidities; the most common number was six. Fig. 6 shows survival stratified by number of comorbidities. The best survival was in patients with four or fewer; survival became progressively worse as that number increased. Nevertheless, we would not say that CABG was futile, even for patients with more than 10 comorbidities who had only a 50% 5-year survival.
|
|
Hear the tale of two surgeons. Fig. 8 shows hospital mortality for surgeons 097 and 125, practicing at a community hospital that Dr John Kirklin and I were asked to review. Notice that surgeon 097 operated on progressively lower risk patients, demonstrated by reduction of expected mortality. This did not occur for surgeon 125 until 1989, when his patient mix changed abruptly. Now notice that surgeon 097's observed mortality progressively declined from slightly greater than expected to about half of expected by 1989. Surgeon 125's observed mortality was also lower in 1989, but it remained several times higher than expected, even for his low-risk patient mix. The lesson here is that case skimming without neutralizing risk factors is not the answer to reducing surgical failures! It only places low-risk patients at elevated danger.
|
Are there tools we could use to facilitate deciding what is best for the patient? Yes! Breathe life into those otherwise static risk factor equations locked in our journals! Make them useful as strategic decision aids for doctors and patients.
Consider a 59-year-old man with severe ischemic cardiomyopathy, previous CABG, and mitral valve repair. He has an implantable cardioverter-defibrillator for intractable ventricular tachycardia. Does he have an indication for heart transplant?
At present, handwritten patient summary sheets, with a huge amount of data, are used in transplant conference to determine if there is an indication. Integrating this complex information is not something our brains are well equipped to do [18]. From what I have told you about this man so far, could you tell me in quantitative terms what his chances are with optimal medical management compared to various alternatives for treating ischemic cardiomyopathy? What is needed is a strategic decision aid to extend our cognitive abilities by integrating these complex data. This aid must be user friendly, graphical, patient specific, and ‘just in time.’
Once again, at Cleveland Clinic we have generated an interactive, web-based decision aid to graph expected time-related survival after transplant given the patient's condition at listing. We intend to display personalized predicted outcome for all options. These should help the heart failure team determine what is best for the patient.
Selecting what is best is a genuine risk-neutralizing strategy, the implementation of which extends far beyond published general guidelines. It is personalized strategic decision-making [19].
3.2 Imponderables
Neutralizing imponderables requires a special environment to take research and innovation from bench and engineering laboratories to bedside in a timely and practical fashion. There needs to be a broad commitment to the process, and awareness that there may be initial failures.
Consider the climate surrounding the first series of patients whose hearts were repaired using a pump-oxygenator at Mayo Clinic [20]. There was buy-in from multiple disciplines: Jesse Edwards in Pathology, Earl Wood and Jeremy Swan in Physiology, Howard Burchell in Cardiology, Jim DuShane in Pediatric Cardiology, Robert Patrick in Anesthesiology, David Donald in Research, and Richard Jones in Mechanical Engineering. Once begun, it was predetermined that there would be a series of at least five patients, then eight, no matter what their outcome. Although four of those first eight patients died, the intense support of multiple parties, as opposed to circling vultures, ushered in one of the most remarkable eras of medicine.
3.3 Treatment
In treatment, opportunities for neutralizing risk factors exist on multiple planes: accuracy of preoperative evaluation, the operative procedure itself, postoperative care, supervision by senior surgeons, education of patient and family, and communication among caregivers. Let me just touch on the obvious. Surgeons and their patients are aware of three essential ingredients of surgical success: skill, rules, and knowledge [18]. There is no shortcut to practice and breadth of experience for instilling surgical skills and rules until they become absolutely automatic. If you or I needed surgery, we would want the most skillful surgeon and a surgical environment organized to minimize inevitable errors, detect ones that do occur, and compensate for them to prevent injury.
Dr Robert Karp wrote about the ‘Culture of Clarity’ that epitomized Dr Kirklin's operating room [21]. It included
Let me illustrate with a common experience faced by parents of teenagers. A decade ago, my son Bill was turning 16. We set a budget for a car. He chose a vintage 1987 Porsche 944-S. Have any of you had the terrifying experience of teaching your son or daughter to drive? It may be even more terrifying for them!
Bill had watched my wife and me drive for years, so he confidently slipped behind the wheel that first time. Piece of cake! The car lurched forward. He jammed on the brake and turned to me astonished! It was not a piece of cake! Within 15 min, he was exhausted. When he first soloed in the Porsche, we heard the car in the distance trying to make it up our hill. Many changes of gear. Lots of noise. Then silence. A few minutes later he walked up to the house and announced, ‘I’ll never be able to drive up that hill!’
Of course, within a few weeks he learned this skill and, indeed, could climb the hill, negotiate curves, and manage the gears with ease—without having to think about it.
Automaticity is key to this transformation. How can forming eye–hand synaptic connections be facilitated to accelerate automaticity of surgical skills? If in part the answer is innate organization of the brain, then there is a place for intense testing of surgical candidates for eye–hand coordination, visual–spatial orientation, and other attributes of the best surgeons. Our training programs, like our educational programs, may give trainees too much benefit of the doubt, promoting individuals who struggle and probably need to be counseled about a different line of medicine for which they are more suited. Surgery also needs simulators, as difficult to develop as they may be, so that training and retraining, certification and recertification, are up to the standards of the pilots of jetBlue Flight 292, who successfully and spectacularly landed their crippled jetliner at LAX in September 2005.
A key aspect of training, and of surgical practice, is not only skills and rules to handle the routine but also knowledge to manage the unexpected but not unanticipated. Recently, Dr Eric Roselli of Cleveland Clinic examined some of the ingredients that lessened the risk of reoperative CABG in the 1990s. One of the factors was better management of adverse events at reoperation. He examined 961 reoperations performed between July 2003 and July 2004 (Roselli EE, Pettersson BG. Personal communication, June 2005). During that time, 84 adverse events occurred in 70 patients, with graft injury, heart injury, and great vessel injury leading the count. Nearly half the events occurred during pre-pump dissection, and another quarter during on-pump dissection.
By this time, there were clearly articulated preventive and compensating strategies for managing rare events such as these. Preventive strategies included
Compensatory strategies included
A number of other strategies are anticipated, including alternative incisions, off-pump surgery, and even hybrid surgical and percutaneous procedures.
Dr Roselli then examined how effectively the surgeon used these preventive and compensatory measures. Of the 84 adverse events, lapses in preventative strategies accounted for 44. In contrast, effective compensatory measures were used in 75 of the 84 events. Of the 70 patients experiencing an adverse event, 81% were rescued from stroke, myocardial infarction, or death. However, despite heroic rescues, as a group, patients experiencing adverse events had worse outcomes than 891 patients not having an adverse event. He concluded that eight patients suffered stroke, myocardial infarction, or death as a result of failure to rescue.
The clinical inferences for neutralization are apparent. Preventive strategies should be practiced in all reoperations to reduce occurrence of adverse events. When they occur, surgeons must be well prepared with a rich repertoire of well-rehearsed compensatory strategies to rescue patients from surgical failure.
3.4 Institutional factors
Time does not permit going into depth about institutional risk factors that play a behind-the-scenes role in either increasing or decreasing surgical failures. Often in our study of high-risk community hospitals, we found major problems at the institutional level, including dysfunctional blood banks, anesthesiologists flitting from hernia repair to a complex tracheal reconstruction, departments working competitively rather than cooperatively, and unstable staffing. The result was surgical failure.
3.5 Societal factors
Similarly, around the world, societies are reevaluating health care delivery. What is the best model? How is it best paid for? How is equitable access guaranteed? What should the priorities be?
These are neither trivial nor easily solved issues [22]. I want only to reemphasize Dr Aberg's assessment of the importance of our professional societies, which foster development of training and retraining programs and certification standards for postgraduate education, disseminate knowledge in meetings like this, and facilitate inter-institutional cooperation at regional, national, and international levels. All of these help neutralize risk of surgical failure on a societal level.
4. Limitations of thinking beyond the risk factors
In closing, I need to acknowledge some assumptions and limitations of thinking beyond the risk factors. Implementing new knowledge implies improving or introducing process. All of us assume there is a strong linkage between process and patient outcome [23]. However, there is evidence of process–outcome disconnect [24,25].
Patients’ outcome and their consumer-oriented expectations are not tightly linked. They will pick a high-risk hospital close to home over a regional medical center to which they must travel. They will pick a procedure that has known inferior long-term outcomes over one that requires invasive surgery.
On the treatment level, when we introduced the incremental risk factor concept many years ago, Dr Kirklin and I stated that risk factors for surgical failure reflected complexity of the procedure. But complexity and outcome may now be largely dissociated. Drs B. Gösta Pettersson and Derlis Martino recently examined nearly 100,000 operations consisting of 1 or more of 11 stand-alone or 3 add-on components (personal communication, June 2005). They concentrated on 38 unique combinations for which there were at least 50 instances and found that the more complex the operation, represented by number of components, the higher was hospital mortality (Fig. 9 ). However, the relation of these 38 combinations to hospital mortality explained only a fraction of the variability (Fig. 10 ). The ROC area of 0.65 is not much above the chance level of 0.5. In contrast, when just patient factors were analyzed without considering the operation at all, there was a strong relation with hospital mortality (C = 0.81). Thus, preoperative patient factors now dominate predictors of hospital mortality, with little explanation of risk coming from complexity of the operation!
|
|
Other disconnects include the glacial speed with which new knowledge is translated into practical processes that can make a difference at the bedside. The public expects instant availability of new therapies, but is unforgiving when they introduce side effects. Perhaps most discouraging is that what we know about better long-term outcomes for the population from primary prevention is nearly impossible to sell and, therefore, to pay for.
5. Conclusions
In conclusion, risk factors for surgical failure exist on multiple levels, from patient to society. Directly neutralizing them may be possible by programmatically directing research and innovation and by putting into practice what we already know. In the meantime, surgical failures can be reduced by the indirect effect of general improvements in the process and practice of surgery. However, be aware that risk factors are associations, not causes. This leads to non-ignorable and counterintuitive process–outcome disconnect. Who knows? Perhaps John Graunt was just lucky that fleeing foul air was associated with halting Black Death! However, I have demonstrated to you that deliberate, creative thinking beyond risk factors, to how they might be neutralized, has often been spectacularly successful.
Footnotes
Presented at the joint 19th Annual Meeting of the European Association for Cardio-thoracic Surgery and the 13th Annual Meeting of the European Society of Thoracic Surgeons, Barcelona, Spain, September 25–28, 2005.
References
This article has been cited by other articles:
|
|
 |
|
  A. R. Kamdar, T. A. Meadows, E. E. Roselli, E. Z. Gorodeski, R. J. Curtin, J. F. Sabik, P. Schoenhagen, R. D. White, B. W. Lytle, S. D. Flamm, et al. Multidetector Computed Tomographic Angiography in Planning of Reoperative Cardiothoracic Surgery Ann. Thorac. Surg., April 1, 2008; 85(4): 1239 - 1245. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |
