892

Section III - Anesthesia management

893

Chapter 24 - Risk of Anesthesia

Lee A. Fleisher

Much attention has been focused on the risks of anesthesia—surgery-related risks to the patient and occupational risks to the anesthesiologist. From the patient's perspective, it is important to provide an accurate assessment of the probability of complications and to study perioperative morbidity and mortality as a means of quality assurance to improve outcome. Numerous studies have attempted to define the probability of morbidity and mortality, with wide variability of results. Risk indices have been developed to identify patients who have a higher probability of developing complications, and it has become evident that individual genetic makeup can affect outcome. In all of these areas, as well as in clinical discussions with patients, the key issue is what risk are we attempting to define.

For the risk of anesthesia, there are multiple factors that enter into the equation. A myopic perspective could include only the morbidity and mortality that occurs intraoperatively. From a quality-assurance perspective, any death within 48 hours after use of an anesthetic is evaluated for potential relevancy. Other investigators have evaluated 30-day morbidity and mortality as part of their estimation of anesthesia-related and surgery-related risks. With respect to ambulatory surgery, 30-day outcomes may be too long, and complications within the first 7 days may be relevant. For patients with cardiac disease, myocardial infarction or cardiac enzyme release in the perioperative period can have implications for months to years.^{[1] [2]} It is important to define the time perspective when evaluating the studies ( Table 24-1 ).

It is also important to differentiate the potential risk that is solely attributable to administration of anesthesia from the risks that anesthesia may modify. From a quality-assurance perspective, a patient with coronary artery disease who has a perioperative myocardial infarction would have the primary cause of morbidity assigned to the disease, but administration of perioperative β-blockers might have prevented the adverse outcome. Anesthesiologists must view perioperative organ protection as part of their goals to provide the highest quality of care and reduce perioperative risk.^[3] Another example is the use of regional anesthesia to reduce graft thrombosis for patients undergoing infrainguinal arterial reconstruction.^{[4] [5] [6]} One outcome that has benefited from regional anesthesia is graft thrombosis and the need for reoperation or amputation. However, the benefit from regional anesthesia was found in only two studies; the third study had a much lower rate of graft thrombosis, suggesting that the benefit may manifest only

894

Traditionally, anesthesiologists and investigators have focused on issues of death and major morbidity such as myocardial infarction, pneumonia, and renal failure. It is becoming increasingly important to include outcomes that affect economic issues, quality of life, and satisfaction for the patient ( Table 24-2 ). For example, nausea and vomiting that delays or prevents discharge to home is important from the perspective of economics and quality of life. Readmission to the hospital after outpatient surgery is an important component of outcome studies. ^[7] Overall satisfaction with care is increasingly included as an outcome.^{[8] [9] [10]} Further research is required to assess the influence of anesthesia care for nonmorbid outcomes and the importance of these outcomes from a patient-oriented perspective.

Many studies have also looked at what some investigators describe as surrogate end points.^[11] For example, nausea and vomiting that do not require treatment or do not delay discharge do not have the same impact as episodes that do affect these outcomes. Myocardial ischemia on the electrocardiogram is another example of a surrogate outcome that may not lead directly to overt morbidity.^[12] In determining the risk of anesthesia, it is important to define the outcome of greatest significance.

Perioperative risk is multifactorial and depends on the interaction of anesthesia-, patient-, and surgery-specific

TABLE 24-2 -- Examples of common outcome measures

Outcome	Example
Mortality
Morbidity
Major	Myocardial infarction
	Pneumonia
	Pulmonary embolism
Minor	Nausea
	Vomiting
	Readmission
Patient satisfaction
Quality of life

factors ( Fig. 24-1 ). With respect to anesthesia, the effects of the agents and the skills of the practitioner are important. Similarly, the surgeon's skills and the surgical procedure itself affect perioperative risk. From the patient's perspective, the question remains whether coexisting disease raises the probability of complications to a level such that the benefit of the surgery is outweighed by the risk. Anesthesiologists frequently have focused on perioperative risk, but the patient is most concerned with management of the disease process. As the specialty focuses on its role in the 21st century, it is important to acknowledge the patient's perspective and desire to undergo procedures that prolong life or improve quality of life. Increasing the ability of the anesthesiologist to affect the decision-making process and the overall risk is the future challenge.

Given these various goals, perspectives, and influences, this chapter attempts to define the current state of knowledge in this area. With the increasing interest in evidence-based medicine, it is important to define what is known and not known. Because it would be unethical to compare operations performed with the use of an "ideal anesthetic" with those using a less ideal or no anesthetic, virtually all studies looking at the factors that contribute to perioperative mortality involve evaluations of large cohorts of patients. However, since the 1990s there have been several well-designed, randomized clinical trials to evaluate different anesthesia regimens, which has advanced the field. This chapter also reviews the literature, with an emphasis on the strength of the evidence for the conclusions.