Magnitude of the problem

Throughout the last few decades, non-cardiac surgery has made substantial advances in treating diseases and improving patient quality of life, and as a result, the number of patients undergoing non-cardiac surgery is growing. A recent study used surgical data from 56 countries around the world to determine that globally there are over 230 million major surgical procedures undertaken annually.¹ The fact that cardiac and pediatric surgery only account for a minority of major surgical cases suggests that over 200 million adults undergo major non -cardiac surgery annually.

Non-cardiac surgery is associated with an increased risk of major vascular complications (i.e. vascular death, non-fatal myocardial infarction, non-fatal cardiac arrest, and non-fatal stroke). A recent review found only one study, by Lee and colleagues,² that evaluated the incidence of major perioperative vascular complications in a prospective cohort that fulfilled the following criteria: more than 300 patients, the patients were relatively unselected (i.e. not restricted to patients referred to internal medicine or to patients with or at high risk of coronary artery disease), not restricted to a specific type of surgery (e.g. orthopedics), and required patients to have at least one measurement of a cardiac biomarker or enzyme after surgery.³ This study suggests that major perioperative vascular events occur in 1.4% (95% confidence interval (CI) 1.0 – 1.8%) of adults undergoing elective non-cardiac surgery. Conservative estimates suggest that at least half of the 200 million adults undergoing non-cardiac surgery are in an at-risk age group.⁴ These data together with the data from the Lee study suggests that worldwide, 1 – 1.8 million adults suffer a major perioperative vascular complication annually.

There is concern, however, that this is a substantial underestimation of the actual incidence. The recent Peri-Operative ISchemic Evaluation (POISE) trial included 8351 patients from 190 hospitals in 23 countries in a randomized controlled trial comparing the effects of a beta-blocker relative to placebo among patients undergoing non -cardiac surgery.⁵ The incidence of major perioperative vascular complications in the POISE trial was more than three times higher than what was predicted by the Revised Cardiac Risk Index, which was developed from the study by Lee and colleagues. This suggests the possibility that the current incidence of major perioperative vascular complications is substantially higher than the estimate from the Lee study. Several limitations associated with the Lee study support this position. Lee et al did not include stroke as an outcome, and excluded patients undergoing emergency surgery.² Emergent cases represent about 10% of non-cardiac surgeries⁶ and patients undergoing emergency surgery are at higher risk of major perioperative vascular events than patients undergoing elective surgery (odds ratio (OR) 2.6, 95% CI 1.2 –5.6).⁷ The data from Lee and colleagues are over 15 years old. Considering that patients with coronary artery disease are living longer and there has been a practice pattern shift towards advanced care for the elderly (including surgery), this suggests that older patients with high burdens of coronary artery disease are now surviving to develop other conditions that require an operation and these elderly patients are undergoing surgery.

These limitations support the finding in the POISE trial, and this suggests that the current worldwide incidence of adults suffering a major perioperative vascular complication in the first 30 days after surgery is probably in the range of 3 – 5.4 million annually. This is in the range of the global incidence of new patients acquiring human immunodeficiency virus (HIV) annually,⁸ and identifies major perioperative vascular complications as a similarly common and important public health problem.

Preoperative risk prediction

Four clinical models developed to predict perioperative cardiac events fulfill the following criteria: assessed in >300 patients, validated at least within the original study, and not restricted to a specific type of non-cardiac surgery (e.g. vascular surgery).^2,7,9,10 Table 5.1 presents information on the studies that developed these clinical models and Table 5.2 presents their scoring systems. All studies had a small to moderate number of events and were underpowered to assess the independent effect of each variable given the number of variables assessed. These clinical models were designed to predict cardiac outcomes of varying importance, and no model included stroke. Table 5.3 presents the studies that have evaluated the predictors of stroke in patients undergoing non-cardiac surgery.^11–15 Although there are few studies and most observed few events, they do suggest overlap between the predictors of stroke and major cardiac events in patients undergoing non-cardiac surgery. The study with the most number of perioperative strokes (i.e. 61 strokes) suggests that perioperative stroke is a serious outcome (i.e. 18% died and 31% were care dependent at discharge).¹⁵

The Revised Cardiac Risk Index is the simplest of the models to use and is more predictive than the Original Cardiac Risk Index.^2,7 Of the two Bayesian models, the Veterans Affairs Model is simpler to use and the one study that compared these two models suggested the Veterans Affairs Model was as predictive as the Modified Cardiac Risk Index.⁷ Although several studies have compared the predictive accuracy of the generic and Bayesian risk models,^2,7,9,16,17 only two have used pretest probabilities in the Bayesian models based upon contemporary data from the hospitals included in the studies.^7,9 The most recent of these two studies demonstrated superior prediction capabilities of the Bayesian risk models.⁷

Table 5.1 Studies that have developed clinical models to predict cardiac events in patients undergoing non-cardiac surgery

MI, myocardial infarction; CHF, congestive heart failure.

Table 5.2 Clinical models to predict cardiac events in patients undergoing non-cardiac surgery

abn rhythm, rhythm other than sinus or premature atrial contractions on last preoperative electrocardiogram; APE, alveolar pulmonary edema; CCSC, Canadian Cardiovascular Society Class; CHF, congestive heart failure; Cr, creatinine; CVA, stroke or transient ischemic attack; IHD, ischemic heart disease; JVD, jugular vein distension; mo, months; poor status, PO ₂ < 60, PCO ₂ > 50 mmHg, K < 3.0, HCO3 < 20 meq/L, blood urea nitrogen > 50, creatinine > 3.0 mg/dL, abnormal serum glutamic oxalacetic transaminase, signs of chronic liver disease, or patient bedridden from non-cardiac causes; Pts, points; PVCs, premature ventricular contractions; U/A, unstable angina; vasc, vascular; VAS, valvular aortic stenosis; low-risk Sx, low-risk surgery (e.g. ophthalmology, maxillofacial, plastic, low-risk orthopedic, urologic, general surgery and non-thoracotomy thoracic procedures); Interm-risk Sx, intermediate risk surgery (e.g. neurosurgery, ENT surgery, major orthopedic surgery); high-risk Sx, high-risk surgery (e.g. intraperitoneal, intrathoracic, or aortic surgery); misc vasc Sx, miscellaneous vascular surgery (e.g. all lower extremity amputations, arteriovenous access procedures).

Despite this finding, the current predictive accuracy of the Modified Cardiac Risk Index and the Veterans Affairs Model is uncertain because there is no high-quality study that has established contemporary complication rates for individual surgeries or groups of comparable surgeries, and it is unknown if contemporary complication rates at one institution are applicable to another institution. Further, it is unclear whether the Bayesian models would demonstrate superior prediction capabilities if the generic risk models included more specific surgeries or groups of comparable surgeries as risk factors, because it is likely that the type of surgery a patient undergoes influences the probability of a vascular complication.

The Vascular events In non-cardiac Surgery patIents cOhort evaluatioN (VISION) Study is a prospective cohort study of 40 000 patients undergoing non-cardiac surgery in several countries around the world (ClinicalTrials.gov, number NCT00512109). A primary objective of this study is to develop a current clinical risk estimation model for predicting major perioperative vascular complications. Until further studies like the VISION Study are complete, physicians need a practical clinical risk index. The Revised Cardiac Risk Index from the Lee study is the simplest to use in clinical practice, and several studies have validated the factors included in the index as independent predictors of major perioperative vascular complications.¹⁸ The Revised Cardiac Risk Index consists of six equally weighted risk factors: high-risk surgery (intraperitoneal, intrathoracic or suprainguinal vascular surgery), history of ischemic heart disease, history of congestive heart failure, history of cerebrovascular disease (i.e. stroke or transient ischemic attack), use of insulin therapy for diabetes, and a preoperative serum creatinine >175μmol/L (>2.0mg/dL). Table 5.4 presents data from the original study and demonstrates the percentage risk of patients suffering a major perioperative cardiac event (i.e. cardiac death, non-fatal myocardial infarction, and non-fatal cardiac arrest) and the corresponding 95% confidence intervals, based on the number of risk factors met. As stated above, the observed event rate in the recent large international POISE trial demonstrated an observed event rate that was three times higher than that predicted by the Revised Cardiac Risk Index. Therefore, physicians may want to double or triple the estimates from Table 5.4 until further data are available.

Table 5.3 Predictors of stroke in patients undergoing non-cardiac surgery

AOR, adjusted odds ratio; CVD, cerebrovascular disease; COPD, chronic obstructive pulmonary disease; PVD, peripheral vascular disease; MAP, mean arterial pressure; TIA, transient ischemic attack; NR, not reported; neurologic symptoms, optic migraine, transient ischemic attack, and stroke.

Non-invasive cardiac stress testing

Although clinical risk indices help to identify patients at risk of a major vascular complication, they underestimate risk in a substantial proportion of patients.^3,17 This likely occurs because for a prolonged period of time prior to undergoing surgery, many patients have limited mobility (e.g. patients with arthritis, peripheral vascular disease, cancer). Some of these patients have underlying cardiac disease but are not active enough to experience symptoms. Although it is likely that researchers will develop more accurate clinical risk indices, it is also likely that a substantial proportion of patients undergoing non-cardiac surgery (e.g. patients with limited mobility, diabetes) will require further evaluation beyond clinical symptoms to optimize their risk assessments.

To address this problem researchers have assessed the prognostic capabilities of non-invasive cardiac stress tests (e.g. stress echocardiography, nuclear scintigraphy imaging), prior to non-cardiac surgery.³ A recent meta-analysis evaluating these two tests demonstrated that they have only moderate negative likelihood ratios (stress echocardiography 0.23 and stress perfusion imaging 0.44), and more than a third of the patients who suffered a major perioperative cardiovascular event had a negative test result.¹⁹ The studies that directly compared these tests suggested there was no difference between the two tests regarding their positive likelihood ratios. When either of these tests had a moderate-to-large defect it was associated with a positive likelihood ratio of 8; however, this test result was only present in a small proportion of patients (i.e. < 15% of the patients undergoing one of these tests).

Table 5.4 Perioperative cardiac risk estimation based upon predictors in the R evised C ardiac risk Index

This table is reproduced with permission from Devereaux et al.³ Risk factors = high-risk surgery (intraperitoneal, intrathoracic or suprainguinal vascular surgery), history of ischemic heart disease (defined as a history of myocardial infarction, positive exercise test, current complaint of ischemic chest pain or nitrate use, or ECG with pathologic Q waves; patients with prior coronary bypass surgery or angioplasty were included only if they had such findings after their procedure), history of congestive heart failure (defined as a history of heart failure, pulmonary edema or paroxysmal nocturnal dyspnea; an S3 gallop or bilateral rales on physical examination, or a chest radiograph showing pulmonary vascular resistance), history of cerebrovascular disease (i.e. stroke or transient ischemic attack), use of insulin therapy for diabetes, and a preoperative serum creatinine > 175 μ mol/L (> 2.0 mg/dL).

Major perioperative cardiac event = cardiac death, non-fatal myocardial infarction or non-fatal cardiac arrest; note that this table does not include postoperative cardiogenic pulmonary edema and complete heart block, which were included as outcomes in the Lee index.

These data likely represent a best-case scenario for these tests because the results report the direct association between the non-invasive cardiac stress test and the risk of a major perioperative cardiovascular outcome. Because of the cost and time associated with these tests, the more relevant data are whether these non-invasive cardiac stress tests provide additional predictive value, beyond clinical variables, for the occurrence of major perioperative cardiovascular events. Most of the studies have not assessed whether these non-invasive cardiac stress tests provide independent prognostic information. The studies that have undertaken multivariable analysis are unreliable because they did not include all the known independent clinical variables or the analysis had too few events for the number of variables assessed.^20–24 The results and the limitations of these data leave considerable uncertainty as to the role of non-invasive cardiac stress testing prior to non-cardiac surgery.

Preoperative BNP and NT-pro BNP measurement

Ventricular myocytes secrete brain natriuretic peptide (BNP), a prohormone, and its inactive cleavage product N-terminal fragment of the prohormone (NT-proBNP) into the blood in response to myocardial stretch and ischemia.^25,26 Plasma BNP and NT-proBNP are powerful predictors of death and major adverse cardiovascular events in patients with stable coronary artery disease, acute coronary syndromes, and congestive heart failure.^27–31 Table 5.5 reports six studies that have evaluated the independent association between a preoperative BNP or NT-proBNP measurement and major perioperative cardiovascular events.^32–37 All studies suggest that an elevated preoperative BNP or NT-proBNP measurement is a strong independent predictor of a major perioperative cardiovascular event. Although there are limitations to these studies regarding the number of events, variations in the outcomes and thresholds evaluated, and the width of the confidence intervals, these results are encouraging. A large substudy (10 000 patients) of the VISION Study currently under way is evaluating the prognostic capabilities of preoperative NT-proBNP and will provide further insights into whether NT-proBNP, a relatively cheap and easily accessible test compared to the non-invasive cardiac stress tests, can enhance perioperative clinical risk predictions.

Table 5.5 Adjusted association between pre-operative BNP/NT-p ro BNP level and perioperative cardiovascular event

MI, myocardial infarction; pulm., pulmonary; VT, ventricular tachycardia; ACS, acute coronary syndrome.

* Authors employed six age-and sex-dependent thresholds;

† resulting in hemodynamic compromise or requiring intervention.

Perioperative myocardial infarctions

These results are similar to the findings in the recent POISE trial, which also monitored cardiac biomarkers or enzymes in all patients for the first 3 days after surgery. Box 5.1 presents the definition used for myocardial infarction in the POISE trial, in which only 35% of the patients suffering a perioperative myocardial infarction had ischemic symptoms.⁵ The fact that approximately 75% of these myocardial infarctions occurred within the first 48 hours after surgery helps to explain why so many patients may not have experienced ischemic symptoms (i.e. this is a period when the majority of patients receive high-dose analgesic medication to blunt surgical discomfort). These data suggest perioperative myocardial infarctions present differently from the majority of myocardial infarction in the emergency room and that perioperative myocardial infarctions are at high risk of going undetected if troponin measurements are not monitored for the first few days after surgery. The ongoing VISION Study will provide further insights into this issue in a large unselected population of patients undergoing non-cardiac surgery who are having troponins monitored throughout the first 3 days after surgery.

Do perioperative myocardial infarctions matter and is surveillance justified?

Stay updated, free articles. Join our Telegram channel

Join