A 75-Year-Old Woman With an Abdominal Aortic Aneurysm and Emphysema

DR PARKER: Mrs H is a 75-year-old woman weighing the risks and benefits of proceeding with an elective surgical repair of an abdominal aortic aneurysm. She lives in the greater Boston area with her husband and has her health care insurance through Medicare.

Mrs H's physicians have followed the size and extent of her thoracic and abdominal aortic aneurysms carefully for several years. By angiogram, the aneurysm involved the origin of the celiac and superior mesenteric arteries, down to the level of the renal arteries. An arteriogram in 1992 revealed a diameter of the thoracic section at 5 cm to 6 cm. An ultrasound in 1995 showed an increase to 6.5 cm. The abdominal aortic aneurysm was measured at approximately 5 cm. By 1996, a magnetic resonance angiogram estimated the thoracic aneurysm at 9 cm, and a computed tomographic (CT) scan estimated its size at 8×12 cm. Mrs H remained asymptomatic during this period with regard to the dilated aorta.

Her medical history is extensive and complex. She has severe emphysema requiring long-term supplemental oxygen use. She had a transient ischemic attack and a central retinal artery occlusion with resulting partial blindness. In 1995, she had pneumococcal pneumonia complicated by septic shock, acute renal failure, encephalopathy, and gastrointestinal bleeding. She has had "chest pain" with a negative workup for coronary artery disease.

In 1997, she complained of the acute onset of severe back pain, and suspected correctly a rupture of her aorta. She was rushed to the local hospital where she was initially resuscitated with fluids, diagnosed as having a ruptured thoracic aortic aneurysm, and transferred to a Boston teaching hospital. She survived a repair of the ruptured thoracic aorta and had a lengthy recovery. She is a former smoker. She currently lives at home and still enjoys activities with her husband, children, and grandchildren even though she is limited by her emphysema.

Her current medications include metoprolol tartrate, 50 mg twice daily; aspirin, 325 mg daily; dipyridamole, 75 mg twice daily; and ipratropium bromide, salmeterol xinafoate, and albuterol sulfate inhalers.

Physical examination reveals a thin, tired-appearing woman who is able to walk 60 m with supplemental oxygen. Her respiratory rate is 22/min, her heart rate is 58/min, and her blood pressure is 118/70 mm Hg. Her lungs are clear, and her heart sounds are regular and distant. She has palpable femoral and pedal pulses bilaterally.

Routine laboratory test results are unremarkable. Pulmonary function tests from 1995 revealed a forced expiratory volume in 1 second (FEV₁) of 1.15 L, with an FEV₁ to forced vital capacity ratio of 53%. Diffusing capacity was markedly reduced at 38%. There was no reversibility with bronchodilator. Mrs H most recently had her abdominal aortic aneurysm measured at 6 cm by ultrasound and 4 cm by CT scan.

Mrs H expressed ambivalence about having surgery on the abdominal aortic aneurysm. She and her family continue a dialogue with her physicians about the merits of further "watchful waiting" vs proceeding with surgery. A vascular surgeon reviewed her case.

MRS H: It was about 11:00 at night. I was having a glass of milk and a couple of cookies, standing up at the counter, and all of a sudden I got this pain [in my back]. It was a pain that I never got before. And right away I knew it was the aneurysm. My husband was lying down, and I yelled for him. That's all I remember before the surgery.

I don't want to linger on. I tell my son, don't let me linger on too long. I don't want to put my family through anything. But I was never afraid. And if I had to go through another operation, I think I could. But I would talk with my family first. I would let them help me decide—one way or the other. If they tell me to go for it, I will. If I was still in this condition and if I could still get up, get myself dressed—which is a hassle, but I do it—and if I could still do little things, I would go for it again. But if I was to go into a wheelchair, or a nursing home, being an invalid, then no. That's how I feel.

MRS T: One doctor came out and told us that it was a grave situation. They said [the aneurysm] had gone to 13 cm. The doctor proceeded to tell us that the aneurysm had blown across her chest. They said that she could die within the next few minutes. It might be an hour. That she definitely wouldn't survive if she was not operated on. And he said there was a chance she could be paralyzed, which wasn't very consoling. We looked at my mother-in-law and she was awake. The doctors asked her if she wanted to have this surgery. And we said, "It's up to you, Mom." And she said, "Well, I'm not going to be here if I don't have it." And she signed her own release right there.

I feel the same way that she does [about another operation]. If they felt that she could make it through another surgery and at least live the way she does now, I think she should have it done. But if it's going to mean that she's going to have less quality of life than she does now, I know it would kill her to have less than what she has now.

DR B: Mrs H is a 75-year-old lady that I have followed for several years. She has atherosclerosis. She had a thoracic aneurysm and an abdominal aneurysm. She had a transient ischemic attack, and she has atypical chest pain. But we have never had any objective evidence of coronary ischemia. She had negative thallium stress test results in the past. Currently, she has a significant abdominal aneurysm that measures between 4 cm and 6 cm that is totally asymptomatic.

There is a significant risk that she will die from this abdominal aneurysm rupturing over the next few years. The trick is to gauge the risk of surgical intervention for this woman with severe lung disease and multisystem disease. She still needs time to recoup from this major operation that she had. We should employ careful surveillance of this aneurysm. If it is enlarging to significantly greater than 6 cm, then we would have to seriously consider surgery. I would do that in conjunction with my pulmonary colleagues because the prime determinant of her surgical risk is her emphysema.

Getting a handle on the risk is important because we know she is at risk if we do not do anything to the aneurysm. We know she is at significant risk of rupture over the ensuing years. And having had the experience once, she is putting pressure on us to strongly consider surgery.

What are the general principles guiding preoperative risk assessment and how do they apply to this patient? How do we assess her surgical risk, and specifically her pulmonary and cardiac risk during and after abdominal aortic aneurysm repair? How do we objectively assess Mrs H's risk for rupture of the abdominal aortic aneurysm without surgery? How do we help Mrs H weigh the risks and benefits of surgery vs "watchful waiting"? How does her current quality of life and outlook color the decision-making process? What are your recommendations for Mrs H?

DR GOLDMAN: Preoperative assessment for elective surgery requires careful assessment of the natural history for which the surgery is being considered, the risks of the procedure itself, the possible influence of other comorbid conditions on surgical risk, the expected long-term prognosis after successful surgery, the prognosis associated with any comorbid conditions, and the patient's preferences. In an era of rapid technological advances, evaluation of both the surgical and nonsurgical alternatives must be tempered by an appreciation of ongoing therapeutic innovation and the likelihood of future discoveries. The process becomes further complicated, but hopefully better informed, by interactions among the surgeon, anesthesiologist, primary care physician, and any medical consultants. For Mrs H, information on preoperative assessment and on the likely risks and outcomes of her underlying conditions and surgical options can guide our recommendations.

Mrs H is not the typical patient, as she presents now after successful repair of a thoracic aortic aneurysm. The case for prophylactic surgery for abdominal aortic aneurysms is based on convincing evidence that infrarenal aortic aneurysms of 5 cm to 6 cm rupture at a rate of about 5% to 11% per year,^{1 - 2} and larger aneurysms rupture at substantially higher rates.³ Conversely, for aneurysms below 4 cm in diameter, the annual rupture rate is close to zero.¹ These small aneurysms tend to expand by an average of about 0.2 cm per year, but the rate of growth is highly variable among individuals and, from year to year, in the same patient.^{1 ,4} Ultrasonography and CT do not always give the same estimates of aneurysm diameter. Measurement error may be as great as the expected annual average change in size, and some aneurysms followed by annual screening actually are reported as being smaller on some follow-up examinations.¹ Ultrasonography is the best screening test because of its lower cost, but CT is considered more accurate for measuring and following the size of the aneurysm.

Several different operative approaches to abdominal aortic aneurysms carry operative mortality rates of about 2% to 5%.^{3 ,5 - 6} In the open endoaneurysmorrhaphy approach, the aneurysmal aorta is opened and the patent branch vessels are oversewn from within. A prosthesis is anastomosed to the proximal aorta above the aneurysm and to the normal distal aorta or common iliac arteries below it. In the exclusion technique, the aneurysmal aorta is ligated proximally and distally and left in place; a bypass graft is placed from the proximal to the distal aorta via an open operative approach. In one small randomized series, the exclusion approach was associated with a nonsignificant 2% absolute reduction in mortality and a significant 13% absolute reduction in the rate of postoperative complications compared with the open endoaneurysmorrhaphy approach, but the former technique also carries a higher, though undefined, risk of expansion and even rupture of the excluded aneurysmal sac if any nonligated vessels still supply it with blood.⁷ More recently, stent grafts have been placed into the aneurysmal aorta via the transluminal femoral approach with procedural mortality rates in the 2% to 5% range and an initial success rate of about 80% to 85% for treating the aneurysm.^{8 - 9} The endovascular approach is of potential advantage for patients with severe pulmonary or cardiac disease because it can be performed without the stress of open invasive surgery and sometimes even under regional anesthesia.¹⁰

About two thirds of the deaths from abdominal aortic aneurysm surgery are related to cardiac causes.¹¹ In the past 20 years, multifactorial indices have provided a reasonable estimate of the probability of cardiac complications from major noncardiac surgery.^{12 - 15} The original cardiac risk index,¹² derived from 1001 patients undergoing major noncardiac surgery, uses 9 factors that were independently associated with major cardiac outcomes to divide patients into 4 risk groups. On this index, Mrs H would have 11 points (age >70 years [5 points], poor general medical status [3 points], and aortic surgery [3 points]) and be in the low-risk but not very-low-risk category with an expected major cardiac complication rate in about the 3% to 10% range depending on the planned surgery.^{12 ,16} In published series that, in aggregate, include more than 4000 patients, the index has performed extremely well in populations such as unselected consecutive surgical patients^{15 ,17} and patients who had had previous cardiac surgery.¹⁸ This index was a statistically significant but less accurate predictor in patients referred for medical consultation¹³ or undergoing abdominal aortic surgery.^{19 - 20}

Subsequently, Detsky and colleagues¹³ used 11 clinical findings to develop a modified multifactorial index and tested it in 268 major operations. By this index, Mrs H would have 10 points (age >70 years [5 points] and poor general medical status [5 points]) and be considered at low risk for cardiac complications. The modified index worked slightly, but not significantly, better than the original index in their own patients but was not evaluated in the 2 largest series of more than 3000 patients in which the original index has been tested and performed well.^{15 ,17} Most studies show the 2 indices are roughly equivalent predictors, but in one study of vascular surgery patients, the original index was a statistically significant predictor while the modified index was not.²⁰ In our own more recent study of over 4000 patients undergoing elective surgery of all types, the original index performed significantly better than the modified index (T. H. Lee et al, unpublished data, 1998). Therefore, in my opinion, no data indicate that the modified cardiac risk index predicts better than the original index, and some data support the opposite conclusion. Furthermore, my colleagues and I have developed and tested a newer index (T. H. Lee et al, unpublished data, 1998) that in our evaluation was preferable to either of the old indices for all patients and, in vascular surgery patients, was also preferable to a model recently proposed by L'Italien and colleagues.¹⁴

Indices derived in consecutive, unselected patients tend to underestimate complication rates when they classify vascular surgery patients as being at low risk,^{19 - 20} especially when the vascular disease sufficiently limits exercise capacity to obscure cardiac symptoms that would be precipitated by activity. As a result, numerous studies have evaluated the potential predictive information that could be obtained from preoperative testing with dipyridamole thallium scintigraphy, stress echocardiography, or ambulatory ischemia monitoring.^{21 - 23} The former 2 are currently the most widely used and, in patients referred for preoperative testing, consistently demonstrate impressive predictive values. The data are far less impressive, however, when such tests are used for preoperative screening in all vascular surgery patients.^{16 ,21} For example, the largest single study of preoperative dipyridamole thallium scintigraphy for the evaluation of patients prior to abdominal aortic aneurysm surgery found that the test was not a significant predictor of myocardial infarction or cardiac death.²¹ Nevertheless, Mrs H's history of a negative dipyridamole thallium study result in the past further supports her relatively low cardiac risk.

In the past several years, various approaches have been proposed by the American Heart Association and the American College of Cardiology,²⁴ the American College of Physicians,²⁵ and other individuals¹⁶ for using preoperative indices to provide reasonable estimates of the risk of cardiac complications with noncardiac surgery and showing how these risks can be modified with the selective use of diagnostic tests. These various approaches generally would recommend noninvasive cardiac testing in an individual whose exercise capacity was severely limited by a noncardiac condition, either by the vascular disease itself or, as in Mrs H, by pulmonary disease, to be sure that substantial ischemic cardiac disease was not being missed. Otherwise, based on a substantial volume of data, preoperative testing can be reserved for patients with intermediate risk based on factors from the history, physical examination, and electrocardiogram. Low-risk patients do not need preoperative testing. High-risk patients by clinical criteria also need not be tested to confirm their high-risk status, but they may need angiography to determine whether preoperative coronary revascularization is likely to be beneficial.^{14 ,16 ,24 - 25} Mrs H is in the low-risk category for cardiac complications and would not need preoperative, noninvasive cardiac testing according to existing guidelines (Table 1).^{16 ,24 - 25}

In the Coronary Artery Surgery Study's registry of patients who had stable class 2 angina or were asymptomatic after a myocardial infarction, total mortality after noncardiac surgery was 2.4% in patients whose coronary disease was managed medically and 0.9% if it was managed surgically.²⁶ However, mortality from elective coronary artery bypass surgery was 1.4%, so the overall mortality was similar for medically treated patients (2.4%) and surgically treated patients (2.3%). For patients with more severe coronary disease, however, there appears to be more of a benefit from preoperative coronary bypass surgery.²⁷

Medical therapy is also improving. In a randomized trial, treatment with atenolol did not reduce in-hospital cardiac or noncardiac death, but at 8 months there was a significant decrease in total mortality, which persisted at 2 years.²⁸ More recent data with mivazerol, an α₂-adrenoreceptor agonist with sympatholytic effects, suggest that this drug has antiischemic benefit perioperatively²⁹ and may reduce myocardial infarction and death in vascular surgery patients (M. F. Oliver, unpublished data, 1998).

Why are risk predictors imperfect? In a case series of patients who had preoperative coronary angiograms and then suffered postoperative myocardial infarction or cardiac death, events were more likely in patients with more severe coronary disease.³⁰ However, none of the events were caused by stenoses of 70% to 99%; 8 were related to inadequate collaterals beyond a total occlusion and 6 occurred in the distribution of arteries without significant preoperative obstructions. Preoperative test results and risk evaluation correlate with the burden of coronary disease, which correlates with the risk of future events, and may also reveal a critical stenosis or a total occlusion with inadequate collaterals. These methods, however, cannot predict new plaque rupture and acute thrombosis in a previously nonstenotic coronary segment.

In general, patients with lung disease tolerate intubation and general anesthesia, with occurrence of death from respiratory failure or pneumonia being unusual even when the FEV₁ is less than 50% of predicted.³¹ In a study of 2291 elective abdominal operations, pulmonary complications were independently predicted by an abnormal pulmonary examination, an abnormal chest radiograph, the original cardiac risk index score, and the Charlson comorbidity index score,³² but not by preoperative spirometry. Nevertheless, patients with resting PO₂ below 55 mm Hg or PCO₂ above 45 mm Hg or FEV₁ below 500 mL are probably at increased risk. For elective abdominal aortic aneurysm surgery, a recent report of 246 original patients combined with a meta-analysis of a number of other series found that impaired pulmonary function, defined as chronic obstructive pulmonary disease, emphysema, pulmonary dyspnea, or previous pulmonary surgery, increased the odds of perioperative death by 1.9-fold after adjusting for age, sex, heart disease, and renal function.³³

Even with many risk factors, it is unusual to find patients whose risk of death from common, elective, noncardiac, nonneurologic surgery exceeds about 10%. This understanding emphasizes the important role that patients' preferences play in influencing recommendations and decision making. For severe conditions, patients may be willing to accept a course of action that might decrease longevity because of its potential to improve quality of life.

Mrs H has a history of chest discomfort that was described as atypical angina and had a normal dipyridamole scan. Furthermore, she has recently survived an extraordinary sequence of events after a ruptured thoracic aortic aneurysm without any major cardiac complications. Based on this history and her prior evaluation, no further cardiac workup would be indicated at this time. Her pulmonary status should be reassessed prior to elective surgery to be sure that any potentially reversible abnormalities were corrected, but it is unlikely that preoperative management could reverse the increase in risk.

Long-term survival after abdominal aortic aneurysm repair, even when performed electively, is lower than for the general population matched for age and sex. In one large series, the annual mortality rate among survivors of elective operation for abdominal aortic aneurysm was about 5.5% per year.³⁴ By comparison, the annual mortality rate for patients who survived surgery after rupture of abdominal aortic aneurysm was about 8.5% in the first 4 years.³⁵

Few medical conditions carry a prognosis as serious as that of an abdominal aortic aneurysm. For Mrs H, the most serious problem is advanced pulmonary disease resulting in a moderate obstructive pulmonary ventilatory defect, severe small airways disease, resting hypoxemia, the need for multiple pulmonary medications, and the ability to walk only about 60 m despite continuous supplemental oxygen. In May 1995, her FEV₁ was 1.05 L to 1.15 L without response to bronchodilators with a PO₂ of 54 mm Hg and a PCO₂ of 41 mm Hg. Her forced expiratory flow rates between 25% and 75% of her vital capacity were all less than 27%, which is consistent with severe small airways disease. Her diffusing capacity was only 38% of predicted capacity. In March 1997, on 1.5 L/min of supplemental oxygen, her PO₂ was 64 mm Hg with a PCO₂ of 45 mm Hg. Her current resting respiratory rate is in the mid-20s.

Lung disease of this severity carries an ominous prognosis, with an estimated annual mortality rate of at least 11% per year³⁶ and with about half of the deaths directly related to chronic obstructive pulmonary disease and pneumonia.³⁷ Mrs H's prognosis from pulmonary disease therefore rivals that of a 5-cm to 6-cm abdominal aortic aneurysm from a statistical perspective, although there is no way to judge which is more likely to be the dominant issue in any one individual.

Based on the risks of surgery and the estimated long-term prognosis with and without surgery,^{1 ,5 - 6} prophylactic surgery for infrarenal aortic aneurysms of 5 cm should, on average, lead to an improvement in life expectancy³ that is worth the cost⁶ in good-risk patients at least up to about age 70 years. For aneurysms below 4 cm in diameter, surgery is not indicated. For infrarenal aneurysms of 4 cm to 5 cm, the official recommendation of the Society for Vascular Surgery and the International Society for Cardiovascular Surgery is elective repair.³⁸ Several studies, however, suggest that an appropriate alternative approach to aneurysms between 4 cm and 5 cm is serial ultrasound every 3 to 12 months with surgery reserved for those that grow by more than 1 cm per year, reach 6 cm, or are associated with any symptoms.⁴ However, such a large proportion of patients qualify for surgery in such a short period that waiting may only delay the inevitable. For example, in another study, 74% of patients younger than 69 years who had aneurysms of 4 cm or greater eventually had surgery.³⁹ Furthermore, if cardiac disease or other diseases progress with age, any benefits of waiting may be offset by the increase in surgical risk. Finally, the risk of dying from emergent surgery for a leaking aneurysm is about 50%, not including the approximately 55% of patients whose aneurysms rupture before they make it to surgery.⁶

At average rupture rates of 3.3%, surgery is preferred.^{3 ,6} However, at rupture rates of 0.7% per year or lower, as found in population-based studies in the United States¹ and the United Kingdom,³⁹ watchful waiting is at least as good a strategy and is probably preferred until the aneurysm reaches about 4.7 cm.³

Mrs H's aneurysm has been variously measured to be as large as 6 cm by ultrasonography in the past and as small as 4 cm by CT recently. It is unlikely that the aneurysm has actually regressed in size, but most experts would rely on the CT measurement rather than the ultrasonographic measurement.

Most importantly, however, all of the data that I have cited on the risks of rupture are for infrarenal aortic aneurysms, not suprarenal or thoracoabdominal aneurysms. Since the normal aorta tapers as it moves distally, a 4-cm aneurysm in the suprarenal aorta is not as worrisome as one in the infrarenal aorta. Most experts recommend a threshold of 6 cm to 7 cm, at most, for prophylactic surgery on the thoracic or thoracoabdominal aorta, except in patients with Marfan syndrome, in whom a 5.5-cm threshold is generally recommended.⁴⁰ Of note is that Mrs H's thoracic aneurysm expanded from 5 cm to 6 cm to at least 8 cm during the 3 years before rupture.

Mrs H's aneurysm involves the origins of the celiac and superior mesenteric arteries and would commonly be defined as an extension of a thoracoabdominal aortic aneurysm, for which the operative mortality is 4.5% or higher.⁴¹ Compromise of spinal cord circulation during or after surgery leads to permanent paraparesis or paraplegia in about 3% to 15% of survivors depending on the level of aortic repair.⁴¹

In our own series of 95 patients undergoing abdominal aortic aneurysm repair, patients' preoperative quality of life was not substantially different than those of age- and sex-matched controls.⁴² These patients then experienced substantial declines in physical function, social function, and vitality at 1 month, but at 6 months and 12 months after surgery their health status equaled or exceeded preoperative values. However, among patients with substantial medical comorbidity, health status at 12 months still tended to be worse than preoperative health status, because progression of the comorbid diseases appeared to outweigh improvements related to surgery. For Mrs H, the concern is that progression of the pulmonary disease will lead to a progressive reduction in health status and quality of life independent of any improvements in vascular status.

Additional data on longitudinal changes in health status and quality of life are available from a Study to Understand Prognoses and Preferences for Outcomes and Risks of Treatments (SUPPORT) analysis of 1438 patients, median age 63 years, with any of 9 serious illnesses (including respiratory failure, severe chronic obstructive pulmonary disease, and multiorgan system failure) requiring hospitalization and carrying an estimated 6-month mortality rate of 50%.⁴³ On day 3 of their hospitalization, seriously ill patients reported a mean quality of life of 73%—in other words, they considered living 1 year in their current state of health to be the equivalent of living 8.8 months in excellent health. However, 35% of patients were unwilling to trade any time in their current state of health for shorter life in excellent health. Of note was that patients' ratings of their quality of life exceeded those of family members by about 8 percentage points and those of their physicians by nearly 16 percentage points. By 2 months later, scores improved by an average of about 6 percentage points and were significantly better in the absence of depression, dejection, tension, anxiety, and physical incapacity. Patients who preferred that their future treatments focus on relieving pain had lower reported quality of life (77%) than those who preferred a focus on life extension (92%; P<.001). Similar results were found in an analogous study of 414 hospitalized patients older than 80 years (median, 84 years).⁴⁴

Anxiety and depression reduce patient-reported quality of life even more than many serious medical conditions.⁴⁵ For example, in one population-based study, the mean quality of life associated with depression (70.3%) was lower than that associated with angina (78.6%) or emphysema (75.1%).

Decision making under uncertainty is, at best, an art form. In an era of sneaker commercials, it is tempting to say, "Just Do It," but sometimes it takes more judgment to decide just not to do it. Mrs H would likely survive aneurysm surgery, but she will certainly have a short-term decline in her quality of life, and likely a permanent decline. She told us if her quality of life was any worse, she would not want to have the surgery. Although I wish it were not true, her quality of life is likely to worsen because she has a progressive and severe pulmonary condition. Fixing her aneurysm may temporarily improve her mental quality of life because she will worry less about rupture, but it will not help her physical quality of life or prevent progressive deterioration. It is critical that Mrs H and her family understand these unfortunate but inescapable realities.

I would not recommend surgery for Mrs H until the aneurysm is at least 6 cm by CT and even then would temper such a recommendation if her pulmonary status has worsened. Given her strong feelings about quality of life, advance directives would be critical to be sure that unwanted procedures are not performed if she later becomes seriously ill. Since her depression and anxiety regarding possible future rupture are the main reversible causes of diminished quality of life, my recommendation is aggressive treatment of the anxiety and depression with medications, while following her aneurysm every 6 months and recognizing that she is more likely to die with, rather than from, this aneurysm.

[EDITOR'S NOTE: Mrs H was unable to attend the conference because of illness.]

DR DELBANCO: What do you think about the formal approaches to measuring patient preferences, for example, patient utilities in helping them to weigh their options?

DR GOLDMAN: I've become a believer in more formal assessments of patient preferences and quality of life. When Dr Phillips and I began our participation in the SUPPORT project, we had misgivings about asking critically ill patients batteries of questions about their preferences for cardiopulmonary resuscitation, their quality of life, their deepest religious beliefs, etc. What was striking to us was that patients not only were willing to answer the questions, but often were thankful that someone asked them. Unfortunately, I was not able to speak with Mrs H directly, since she could not be here today. I wanted to hear how she weighs her fear of aortic rupture with her disability from her pulmonary disease. For Mrs H, I was impressed that she said if she had to be in a wheelchair, life wouldn't be worth living.

AN INTERNIST: In assessing preoperative risk in thoracic surgery, functional status is determined by exercise testing and pulmonary functioning. Can you tell me why pulmonary function and disability is absent from the prediction algorithms you described?

DR GOLDMAN: Pulmonary risk becomes a major issue when you are removing part of a lung and the patient's residual lung capacity is critical. In that situation, partial lung function testing can be helpful in assessing whether remaining lung tissue will be adequate for a reasonable quality of life. For non–lung-resection surgery, a reasonable truism is that if you can breathe before the operation, you can breathe after the operation. When lung disease kills people postoperatively, it is usually because they have multiorgan system failure. Pulmonary problems are rarely a major cause of death in large series of nonlung surgeries.

AN INTERNIST: We have been talking about the potential risk of rupture from abdominal aortic aneurysms with relatively small differences in aneurysm size, but no one has taken into account patient size. Is there any reason that has not been done or should be done?

DR GOLDMAN: The issue of correcting aneurysm size for body size certainly makes sense, but it generally has not been done, or at least I was unable to find any information on it. The normal size of the aorta varies slightly by patient size, so it may be a confounder of the likelihood of rupture. Another confounder is measurement variation from test to test, especially with serial ultrasounds. I think there is a general belief in the literature that the ultrasound is the best screening test when you first see someone and wonder whether they have an aneurysm, but the CT scan is preferred for following patients at 6-month or annual intervals because it is more accurate.⁴⁰

AN INTERNIST: We have been talking about depression in patients, often with chronic illness, who are facing serious treatment decisions along with their physicians and caregivers. How aggressive should we be in diagnosing and treating depression before making these serious decisions?

DR GOLDMAN: It is important to understand how much depression is affecting quality of life and try to intervene with a patient like Mrs H. If we were looking at maximizing her quality of life over the next 3 to 5 years, we probably could do more with antidepressants than with any other single intervention. One way to make her feel better mentally would be to repair her aortic aneurysm, so she would not worry it would rupture. Another way would be to find some way to make her worry less about whether it is going to rupture. The practical intervention of antidepressant medications is probably the one most likely to make a difference in how she or her family look back and rate the quality of these last several years of her life.

AN INTERNIST: This case raises a question of how much of the doctor's own uncertainty and anxiety about these important decisions should properly be transferred to the patient. One alternative to this patient's depression might be to have her physician evaluate the options, make a decision, and not to involve her in this, unless the doctor is convinced that this is the right way to go.

DR GOLDMAN: This may be an example of a patient for whom full disclosure and patient participation has not led to a clear decision, but has led to more anxiety and depression. In this era, how can one not tell patients everything and involve them in decision making? Nevertheless, there is a fine line between involving a patient on the one hand and transferring the physician's ambivalence on the other hand. Some patients take advantage of a physician's ambivalence, others suffer from it. I am not proposing a return to the era when we would not tell people they had cancer unless they asked. However, the more we transfer our uncertainty to patients, the more they wonder about who is in charge and how good the physician's advice is. I certainly believe in patient participation and informed decision making. I also believe that it is important for the physicians involved in a case to frame the issues carefully and give the patient a clear set of options, any of which the physicians are ready to follow based on the patient's preferences, or a single, definitive, unified plan when one option is clearly preferable.