Infarcts after surgery

Assessment of patients before operation can reduce the risk

Decisions about the cardiac risk of a surgical procedure may be taken by the surgeon, anaesthetist, cardiologist, referring doctor, or general practitioner. All need to make informed judgments about the perioperative risk, yet despite advances in anaesthesia and surgical techniques the prediction of risk remains an uncertain science.

The leading cause of death after surgery is myocardial infarction.1 Non-fatal myocardial infarction accounts for considerable morbidity2; it may be silent in a quarter of cases, and (not surprisingly) it is associated with a poor outcome. Yet 17 years after Goldman et al developed a clinical risk score in an attempt to predict which patients would develop cardiac complications3 we are not aware of any prospective trial in Britain intended to measure the incidence of perioperative infarction. Trials in North America have reported rates of serious cardiac complications ranging from 1.7% to 40%.4 These data have been used to justify strategies as various as no attempt at predicting risk and routine preoperative coronary angiography in high risk subsets such as patients with peripheral vascular disease.5 Assessment of the extent of the problem in Britain is, we believe, urgent so that decisions may be made on the need for preoperative assessment and how aggressive it should be.

What methods of assessing risk are available? Goldman et al's original scoring system was modified by Detsky and tested by several authors. It remains useful mainly for predicting which patients are at very high risk6—something often apparent from a clinical assessment. An extensive review in 1990 cast doubt on the value of age and stable angina as predictors of risk and suggested that an aggressive approach to perioperative monitoring might keep the risk of reinfarction following recent preoperative myocardial infarction as small as 5.7%.7 The importance of preoperative hypertension was considered to be “controversial,” though there is evidence that poorly controlled or labile hypertension increases risk. The only independent predictors of risk were congestive heart failure, diabetes, peripheral vascular disease, and a rhythm other than sinus on preoperative electrocardiography. The conclusion seems to be that clinical factors are of limited help.

An exercise tolerance test may be unhelpful in patients at potentially high risk of peripheral vascular disease since they may be unable to exercise or reach 85% of the maximum predicted heart rate.8 Routine coronary angiography carried out preoperatively in 1000 patients undergoing elective peripheral vascular surgery found that 37% of those with no clinical suspicion of coronary disease had angiographic evidence of advanced disease.6 Almost a quarter of the patients went on to cardiac surgery, with a mortality of 5.3%; and since this mortality is similar to that resulting from peripheral vascular surgery alone many centres have continued to advocate that preoperative coronary artery bypass should be used only selectively. In the coronary artery surgery study patients were randomised to coronary bypass or medical treatment before their non-cardiac surgery: no difference was seen in perioperative myocardial infarction between the groups.9

More recently, attention has focused on selecting patients for invasive screening and intervention, with dipyridamolethallium scanning the gold standard for non-invasive prediction of risk.10 A reversible defect on thallium scanning has a sensitivity of 80-90% in predicting perioperative cardiac morbidity,11 but the specificity of such scanning is poor and its usefulness as a predictor has been questioned.12 Published reports are conflicting, but, with the exception of a single large study,13 those investigations that found thallium scanning to be of limited value were based on small numbers. Furthermore, semiquantitative methods of thallium scanning may improve its accuracy.14 Nevertheless, access to scanning facilities is limited and obtaining radioisotopes at short notice is often difficult, so for the foreseeable future many patients undergoing major surgery will not have access to this investigation.

Thallium scanning may be the best preoperative investigation we have, but since its availability is limited the search continues for the ideal non-invasive predictor of risk. Recent work has shown that preoperative 24 hour ambulatory electrocardiographic monitoring may be useful, particularly when combined with thallium imaging. In one study a third of patients who had had a reversible defect in the thallium scan and abnormalities in their electrocardiograms developed a postoperative myocardial infarction.15 Continued monitoring in the postoperative period may identify silent ischaemia, which may be a warning of the onset of a clinical cardiac event, with the possibility of intervention. Attention has also focused on measures of left ventricular function rather than coronary ischaemia. Echocardiography has been found to play a part as a predictor of risk, particularly if used to assess the development of abnormalities of wall motion induced by dobutamine,16 but further work is required. Yet another approach is to use suprasternal Doppler, but since this test is notoriously subjective, its main value may be in the continued monitoring of a single patient. Others have used changes in thoracic electrical impedance to measure pump performance non-invasively, but this has not given consistent results.

Our current methods of predicting perioperative cardiac risk are, then, characterised by uncertainty. Surgeons should take more responsibility for defining the risk in their own population of patients; if it is sufficiently high they should adopt a protocol for assessing individual patients. This will require selection for thallium scanning on the basis of clinical risk factors, with the possibility of coronary angiography for those found to have reversible defects on thallium imaging. Further studies are urgently needed to provide a reliable algorithm for non-invasive assessment of operative risk.