Author + information
- Received July 22, 1999
- Revision received November 30, 1999
- Accepted January 13, 2000
- Published online May 1, 2000.
- Mini K Das, MD∗,
- Patricia A Pellikka, MD, FACC∗,*,
- Douglas W Mahoney, MS†,
- Veronique L Roger, MD, FACC∗,
- Jae K Oh, MD, FACC∗,
- Robert B McCully, MD, FACC∗ and
- James B Seward, MD, FACC∗
- ↵*Reprint requests and correspondence:
Dr. Patricia A. Pellikka, Mayo Clinic, 200 First Street SW, Rochester, Minnesota 55905
This study evaluated the incremental value of dobutamine stress echocardiography (DSE) for assessment of cardiac risk before nonvascular surgery.
Limited information exists regarding the preoperative assessment of cardiac risk in patients with known or suspected coronary artery disease who are to undergo nonvascular surgery.
All patients (303 men, 227 women) who underwent DSE before nonvascular surgery and did not sustain an intervening event (coronary revascularization or cardiac event) were studied. Clinical, electrocardiographic and rest and stress echocardiographic variables were evaluated to identify predictors of postoperative cardiac events.
Events occurred in 6% of patients: 1 cardiac death and 31 nonfatal myocardial infarctions. All of these patients had inducible ischemia on DSE (sensitivity 100%, specificity 63%). Multivariate predictors of postoperative events in patients with ischemia were history of congestive heart failure (p = 0.006; odds ratio = 4.66; confidence interval 1.55 to 14.02) and ischemic threshold less than 60% of age-predicted maximal heart rate (p = 0.0001; odds ratio 7.002; confidence interval 2.79 to 17.61). Clinical variables of Eagle’s index identified 21% of patients as low, 68% as intermediate and 11% as high risk preoperatively; the postoperative event rates were 3%, 6% and 14%, respectively. Dobutamine stress echocardiography identified 60% of patients as low (no ischemia), 32% as intermediate (ischemic threshold 60% or more) and 8% as high risk (ischemic threshold < 60%); postoperative event rates were 0%, 9% and 43%, respectively.
In this population of patients with known or suspected coronary artery disease evaluated before nonvascular surgery, DSE had incremental value over clinical, electrocardiographic and rest echocardiographic variables for identifying patients at low, intermediate and high risk for postoperative cardiac events. Ischemia occurring at less than 60% of age-predicted maximal heart rate identified patients at highest risk.
Perioperative cardiac complications such as cardiac death, myocardial infarction, unstable angina, pulmonary edema and serious ventricular arrhythmias are potentially avoidable causes of morbidity and mortality in patients undergoing major vascular surgery (1). Identification of patients at moderate to high risk for perioperative events should lead to modifications in the management of these patients in an attempt to prevent these events. The presence of inducible ischemia during dobutamine stress echocardiography (DSE) has been shown to be predictive of perioperative cardiac events and has become a widely used and accepted method for preoperative assessment of cardiac risk in patients undergoing vascular surgery (2–6).
Less information exists regarding the role of DSE for preoperative risk stratification before nonvascular surgery (6); the likelihood of coexistent severe coronary artery disease in this more general population is presumed to be lower (1,7–10). However, symptoms of coronary artery disease may be absent or blunted in patients limited by orthopedic, pulmonary or other superimposed debility. Clinical evaluation alone, therefore, may be insufficient for risk stratification in patients with such limitations, and noninvasive testing may be of further benefit. The purpose of this study was to determine the incremental value of preoperative assessment of cardiac risk using DSE before nonvascular surgery in patients with known or suspected coronary artery disease.
From January 1990 to December 1995, 544 patients who were unable to exercise underwent DSE for risk stratification before anticipated nonvascular surgery. Fourteen patients with a positive DSE had an intervening event, including coronary revascularization in 13 patients and myocardial infarction in 1 patient, before nonvascular surgery and were excluded from further analysis. The remaining 530 patients underwent nonvascular surgery within six months after the DSE and constituted the study group.
Clinical characteristics of the patients were prospectively recorded at the time of DSE. The Eagle index (11) was calculated for each patient according to the number of the following criteria that were present: age older than 70 years, history of angina pectoris, history of prior myocardial infarction, history of diabetes and history of congestive heart failure. On the basis of the number of positive criteria, patients were classified as low (none), intermediate (one or two) or high (more than two) risk.
The records of the 530 patients were reviewed for the presence of postoperative hard cardiac events, including cardiac death or acute myocardial infarction. Cardiac events were determined by postoperative serial cardiac enzyme values, postoperative electrocardiograms and echocardiographic data, which were obtained in 320 patients (60%), including all patients who experienced angina, dyspnea, heart failure, significant hemodynamic changes, ventricular arrhythmias or atrial arrhythmias requiring treatment.
Dobutamine stress echocardiography was performed according to a previously described protocol using 3-min stages, a peak dosage of 40 μg/kg/min and a target heart rate of 85% of the age-predicted maximum (12). The mean peak dose was 36 ± 9 μg/kg/min. One hundred twenty-three patients (23%) received atropine (mean dose 0.7 ± 0.5 mg, range 0.2 to 3.0 mg). One hundred two patients (19%) received esmolol (0.5 mg/kg) to reverse the effects of the dobutamine. Use of cardiac medications was not discontinued before DSE. A 12-channel electrocardiogram was recorded at baseline and at each stage, and three-channel monitoring of rhythm was performed continuously. The electrocardiogram was positive for ischemia if there was horizontal or downsloping ST-segment depression of 1 mm or more at 80 ms after the J point.
Echocardiographic imaging was performed continuously during the dobutamine infusion. Ejection fraction was determined by visual estimation or by a modified method of Quinones et al. (13). Regional wall motion was assessed semiquantitatively by an experienced level III echocardiographer blinded to clinical information. Interpretation was based on review of both quad-screen images and videotape recordings. Each of 16 segments was scored at rest and at peak stress according to the American Society of Echocardiography Committee guidelines (14). Failure to develop hyperdynamic function or the development of new or worsening wall motion abnormalities, including worsening of a wall motion abnormality that had improved with a low dose of dobutamine, was interpreted as an ischemic response. Wall motion abnormalities that remained fixed with stress were considered to represent infarction. The wall motion score index was calculated at rest and with stress as the sum of the scores divided by the number of visualized segments. The change in the wall motion score index from rest to stress was measured. The number of abnormal segments divided by the number of segments visualized (percentage of abnormal segments) at rest and with stress was recorded. A change in the percentage of abnormal segments from rest to stress also was measured. Ischemic threshold was defined as the heart rate at which new echocardiographic wall motion abnormalities first occurred divided by the age-predicted maximal heart rate (220 − age) (5).
Continuous variables were summarized as a mean ± 1 SD, and group comparisons were based on the Wilcoxon rank sum test. Categorical variables were summarized as a percentage of the group total, and comparisons between groups were based on the chi-square test. Associations between cardiac events and clinical and echocardiographic variables were examined with logistic regression. Modeling of the event of interest consisted of two stages. Stage 1 modeling consisted of univariately significant clinical and resting echocardiographic variables that entered into the model in a stepwise forward selection manner. Entry and retention were set at a significance level of 0.05. Variables from stage 1 modeling were then used as clinical and resting risk factors for stage 2 analysis, in which stress electrocardiographic and echocardiographic variables were entered into the model in a stepwise forward selection manner.
All subjects who had a hard cardiac event also had inducible ischemia noted on echocardiography. The 100% sensitivity of inducible ischemia resulted in an infinite likelihood function, creating a nonestimable model parameter. Therefore, subjects with noninducible ischemia were considered a low-risk group (negative predictive value of 100%), and the two-stage modeling process was then used for subjects who had ischemia to define intermediate- and high-risk subgroups further. Receiver operating characteristic analysis of ischemic threshold defined ischemic threshold less than 60% of age-predicted maximal heart rate as most discriminating.
The mean age of patients was 71 ± 9 years (range 31 to 92 years). There were 303 (57%) men and 227 (43%) women. The clinical characteristics of the study population are described in Table 1. Reasons for being unable to exercise included orthopedic limitation in 312 patients (59%), peripheral vascular disease in 110 (21%), debility in 52 (10%), pulmonary disease in 24 (4%) and other reasons in 32 (6%). The baseline electrocardiogram was normal in 155 patients (29%) and showed previous myocardial infarction in 113 patients (21%), left ventricular hypertrophy in 27 patients (5%) and left bundle branch block in 24 patients (5%).
DSE end points
Target heart rate was achieved in 382 patients (72%). End points for stopping the study included achievement of target heart rate in 336 patients (64%), maximal dobutamine dose in 92 (17%), new regional wall motion abnormalities of at least moderate severity in 39 (7%), hypotension in 16 (3%), ventricular arrhythmias in 15 (3%), intolerance of symptoms in 12 (2.2%), supraventricular arrhythmias in 8 (2%), angina in 5 (1%), electrocardiographic changes in 3 (0.6%) and hypertension in 1 (0.2%). Chest pain occurred in 69 patients (13%) and dyspnea in 27 patients (5%).
Results of DSE
Images were of adequate diagnostic quality in all patients (limited but diagnostic in 102 patients, 19%). Dobutamine stress echocardiography was normal in 201 patients (38%). One hundred fifteen patients (22%) had evidence of prior infarction without detectable ischemia. Two hundred fourteen patients (40%) had new or worsening wall motion abnormalities during DSE, consistent with ischemia: 84 patients (16%) without resting wall motion abnormalities and 130 patients (24%) with evidence of both ischemia and prior infarction. Left ventricular hypertrophy was present in 112 patients (21%). Dobutamine stress echocardiography was positive for ischemia in 66 patients (12%).
Nonvascular surgery in the 530 patients included orthopedic procedures in 264 patients (50%), abdominal in 81 (15%), head and neck in 81 (15%), pelvic in 68 (13%) and thoracic in 36 (7%). Postoperative cardiac events occurred in 32 patients (6%): nonfatal myocardial infarction in 31 and fatal myocardial infarction in 1.
Predictors of postoperative events
Univariate predictors of postoperative cardiac events are shown in Tables 2 and 3 . ⇓⇓ A negative response on DSE, without evidence of ischemia, identified a low-risk group (p = 0.0001; odds ratio [OR] = 0.009; 95% confidence interval [CI] = 0.0005 to 0.145). Inducible ischemia was noted in all patients with a hard event (sensitivity 100%, specificity 63%, positive predictive value 15%, negative predictive value 100%). We studied the 214 patients with an ischemic response on DSE in an attempt to identify intermediate- and high-risk subgroups.
The clinical and resting echocardiographic variables that were significantly associated with a cardiac event in the ischemic group (stage 1 modeling) were a history of congestive heart failure (p = 0.006; chi-square = 7.6; OR = 4.2; CI = 1.5 to 11.5) and low resting heart rate (p = 0.0021; chi-square = 9.4; OR = 1.7; CI = 1.2 to 2.3, per 10 beats/min increment), and the overall model chi-square value was 14.5. From stage 2 modeling, the only stress echocardiographic or electrocardiographic variable that added incrementally to the clinical model was ischemic threshold less than 60% (p = 0.0001; chi-square = 17.1; OR = 7.0; CI = 2.8 to 17.6), and the final model chi-square value was 31.4. After including the ischemic threshold of less than 60% in the model, resting heart rate was no longer significant (p = 0.16). The final model is presented in Table 4.
The mean ischemic threshold was 62% among patients with a cardiac event and 75% in patients without an event (p = 0.0001). There were 17 cardiac events among 40 patients who had an ischemic threshold less than 60% of age-predicted maximal predicted heart rate (positive predictive power 43%, negative predictive power 57%).
Eagle’s index versus DSE
We compared preoperative risk assessment as determined by Eagle’s index with findings of inducible ischemia and the ischemic threshold as determined by DSE to evaluate the incremental value of DSE in preoperative risk stratification before nonvascular surgery. Table 5provides the cross-tabulation of Eagle’s index with DSE results.
Eagle’s index identified 107 patients (21%) at low risk, 363 (68%) at intermediate risk and 57 (11%) at high risk for postoperative hard cardiac events. These patients had event rates of 3%, 6% and 14%, respectively. The DSE result, based on the presence of ischemia and the ischemic threshold, identified 316 (60%) at low risk, 171 (32%) at intermediate risk and 40 (8%) at high risk for postoperative hard cardiac events. These patients had event rates of 0%, 9% and 43%, respectively. For each group based on Eagle’s index, the DSE result further stratified patients’ postoperative risk. Because the cardiac event rate in this population was 6%, an overall risk index combining both the Eagle index and the DSE result could be defined as low, intermediate and high, as indicated in the total columns of Table 5. Of the three patients in the low-risk group of the Eagle index who had an event, DSE result identified one patient at intermediate risk and two patients at high risk. Of the 21 patients in the intermediate-risk group of the Eagle index who had an event, the DSE result identified 8 patients at intermediate risk and 13 at high risk. Thus, the DSE result consistently provided incremental risk assessment at each level of the clinically based (Eagle’s index) risk assessment.
In this study of 530 patients who were unable to exercise and underwent DSE for cardiac risk stratification before nonvascular surgery, the strongest predictors of postoperative cardiac events were ischemia on DSE and a history of congestive heart failure. Furthermore, DSE results could be used to identify patients at low, intermediate and high risk for postoperative hard cardiac events: no events occurred in 316 patients with a negative DSE (low-risk group), 15 (9%) of the 171 patients with ischemia at 60% or more of predicted heart rate had events and 17 (43%) of 40 patients with ischemia at 60% or more of predicted heart rate had events. Ischemia by DSE was superior to clinical variables for predicting cardiac events.
Clinical prediction of perioperative cardiac events
Perioperative cardiac complications are the leading cause of mortality after anesthesia and surgery (15,16). Clinical predictors of perioperative cardiac mortality and morbidity have been characterized by multiple studies (17–21). However, the indices derived from these studies proved to be cumbersome for daily use and are not sensitive for identifying high-risk patients in an intermediate-risk range (11).
More recently, Eagle et al. (22), in a study of 61 patients who underwent vascular surgery, suggested and validated a simplified approach to perioperative clinical risk stratification that divided patients into low-, intermediate- and high-risk groups. This approach was further validated in a subsequent study of 200 patients (23), by other investigators (24,25) and by concordance with coronary angiography (26). With this index, patients at low clinical risk have a low likelihood of severe coronary artery disease and usually require no further evaluation, whereas patients with an unstable clinical status should be considered for coronary angiography. Patients with intermediate or high risk by clinical markers but with a stable clinical status should be considered for further noninvasive testing (11). However, most studies validating the Eagle index were conducted in patients undergoing vascular surgery.
Cardiac risk assessment before nonvascular surgery
Prior studies (6,24,25,27–29) involving cardiac risk assessment before nonvascular surgery have been few and included a small number of patients (all with fewer than 125 patients). These included studies in which patients underwent either vascular or nonvascular surgery and evaluation was not individualized by surgical type. In 100 patients undergoing nonvascular surgery (24), cardiac risk assessment with dipyridamole-thallium was useful in patients who were older (>70 years) or who had a history of congestive heart failure for predicting postoperative unstable angina, congestive heart failure, myocardial infarction or cardiac death. No large studies are available (30) regarding risk assessment in patients before major nonvascular surgical procedures (especially those at intermediate to high clinical risk).
DSE for prediction of perioperative cardiac events
In our study of 530 patients with a stable clinical status before nonvascular surgery, 79% of whom were at intermediate to high risk by Eagle’s index, the postoperative hard cardiac event rate was 6%. This relatively high event rate, which is comparable with the pooled average for large studies involving patients referred to tertiary-care centers for preoperative testing (30), supports the suggestion that further risk stratification is needed in this population. In addition, three events occurred in patients who were at low risk by Eagle’s index, and these patients were more accurately identified by DSE as intermediate or high risk (Table 5).
Ischemia detected on DSE is predictive of postoperative events in patients undergoing peripheral vascular surgery; the positive predictive value is 13% (range 7% to 42%) and the negative predictive value is 99% (range 93% to 100%) (2–5,31,32). The performance of DSE in our population undergoing nonvascular surgery was similar (positive predictive value 15%, negative predictive value 100%). Earlier studies using dipyridamole-thallium scintigraphy reported a positive predictive value ranging from 25% to 50% when soft cardiac events (congestive heart failure, unstable angina) were included and test results were not used in clinical decision making (31,33). More recent reports that reflect the use of test results to prompt revascularization, avoidance of elective surgery or changes in postoperative management have reported an average positive predictive value of 14% (range 5% to 37%) (27,34–36). In a recent meta-analysis, the prognostic values of DSE and nuclear perfusion imaging in patients undergoing nonvascular surgery were similar (31,37).
Utility of ischemic threshold
In our patient population, the average ischemic threshold of 62% among patients with an event was significantly different from the ischemic threshold of 75% in patients without an event (p = 0.0001). These findings are similar to those of Poldermans et al. (5). By receiver operating characteristic analysis of our data, an ischemic threshold less than 60% was most discriminating. Utilizing ischemia and an ischemic threshold less than 60%, DSE had an incremental value over clinical and rest echocardiographic variables for identifying patients who had a hard cardiac event postoperatively.
Prior studies have shown that postoperative ischemic events appear to be related to an exaggerated sympathetic response with substantial increases in heart rate in the postoperative period (16,38) and that the event rate can be reduced with use of medications such as atenolol to reduce heart rate in the postoperative period (38). Thus, it is not surprising that the data from our study underscore the importance of ischemic threshold as a sensitive index of preoperative cardiac risk. Ischemic threshold measured during DSE has previously been shown to be reproducible and to correlate with both the number of stenosed coronary arteries and the left ventricular ejection fraction response to stress (39). It appears to be an important indicator of prognosis and should be measured routinely during DSE. Of note, DSE is the only commonly used pharmacologic stress test that provides this information.
Routine collection of postoperative serial electrocardiographic and cardiac enzyme studies was not performed in all patients, which may have led to an underestimation of cardiac events. Because clinicians used the results of DSE to manage patients, treatment bias was unavoidable (31). Several patients with markedly positive DSE were treated with revascularization and excluded from the study. In addition, modification of postoperative management in patients with a positive DSE test may have contributed to lowering the event rate. These considerations may lower the positive predictive value of the test. Patients referred for preoperative DSE did not represent a consecutive group undergoing nonvascular surgery. Most were in an intermediate- to high-risk group. Nevertheless, cardiac events also occurred in patients at low clinical risk, and DSE was useful for identifying these patients.
In this population of 530 patients who were unable to exercise and were undergoing nonvascular surgery, DSE was superior to clinical variables for identifying patients at risk of cardiac events. Ischemia occurring early at less than 60% of age-predicted maximal heart rate identified patients at highest risk, despite low or intermediate risk categorization by clinical variables, whereas a normal study conferred an excellent outcome (negative predictive value 100%) regardless of risk stratification by clinical variables. Dobutamine stress echocardiography is the only commonly used pharmacologic stress test that can provide the clinician with an ischemic threshold.
- confidence interval
- dobutamine stress echocardiography
- odds ratio
- Received July 22, 1999.
- Revision received November 30, 1999.
- Accepted January 13, 2000.
- American College of Cardiology
- Davila-Roman V.G.,
- Waggoner A.D.,
- Sicard G.A.,
- Geltman E.M.,
- Schechtman K.B.,
- Perez J.E.
- Poldermans D.,
- Fioretti P.M.,
- Forster T.,
- et al.
- Poldermans D.,
- Arnese M.,
- Fioretti P.M.,
- et al.
- Pellikka P.A.,
- Oh J.K.,
- Bailey K.R.,
- Nichols B.A.,
- Monahan K.H.,
- Tajik A.J.
- Quinones M.A.,
- Waggoner A.D.,
- Reduto L.A.,
- et al.
- Schiller N.B.,
- Shah P.M.,
- Crawford M.,
- et al.
- Mangano D.T.,
- Browner W.S.,
- Hollenberg M.,
- London M.J.,
- Tubau J.F.,
- Tateo I.M.
- Mangano D.T.,
- Hollenberg M.,
- Fegert G.,
- et al.
- Paul S.D.,
- Eagle K.A.,
- Kuntz K.M.,
- Young J.R.,
- Hertzer N.R.
- Lette J.,
- Waters D.,
- Cerino M.,
- Picard M.,
- Champagne P.,
- Lapointe J.
- ↵Leppo JA. Preoperative cardiac risk assessment for noncardiac surgery. Am J Cardiol 1995;75:42–51D.
- Brown K.A.,
- Rowen M.
- Shaw L.J.,
- Eagle K.A.,
- Gersh B.J.,
- Miller D.D.
- Panza J.A.,
- Curiel R.V.,
- Laurienzo J.M.,
- Quyyumi A.A.,
- Dilsizian V.