|Home | About | Journals | Submit | Contact Us | Français|
Routine use to prevent perioperative cardiac complications is not evidence based
Globally, about 100 million adults have non-cardiac surgery each year. In the United States, perioperative cardiac complications occur in 0.5-1%, so around one million patients risk cardiac complications and about a quarter will die each year.1 2 Outcomes in Europe are similar to the US.3 Anaesthetists have progressively changed the emphasis on reducing perioperative cardiovascular risk from assessing preoperative coronary artery anatomy to understanding the pathophysiology of perioperative myocardial ischaemia. Despite efforts to identify risk factors for perioperative myocardial ischaemia and potential therapeutic options in the perioperative period, the benefit of giving β blockers and statins at this time remains unclear.2 4 5
Since the early studies that incorrectly attributed survival benefits to perioperative treatment with β blockers,6 rigorous meta-analysis confirmed the need for a large multicentre randomised placebo controlled trial.5 Since then, 1520 patients have been randomised to three studies that have shown no benefit from perioperative metoprolol.7 8 9
The diabetic postoperative mortality and morbidity study from Denmark recruited 921 patients and found that metoprolol had no benefit in patients with diabetes who were β blocker naive with respect to death, myocardial infarction, unstable angina, or congestive heart failure 30 days after surgery.7
The perioperative β blockade study in the United Kingdom randomised 103 patients undergoing infrarenal vascular surgery and found that perioperative metoprolol did not reduce cardiovascular events at 30 days. Events included all cause mortality, myocardial infarction, unstable angina, ventricular tachycardia, and stroke.9
The metoprolol after vascular surgery study randomised 496 vascular surgery patients and also reported no benefit from perioperative metoprolol in reducing postoperative cardiac events at 30 days and six months.8 These three studies of two groups of patients at moderately high risk of perioperative cardiac complications or death (patients with diabetes and patients with vascular disease), undergoing moderate and high risk surgery, provide no strong evidence that treatment with β blockers in the perioperative period confers any benefit. However, all three studies document a strong association of β blockade with an increased risk of bradycardia and hypotension that will require treatment.7 8 9 The results of these studies have been summarised and coupled with a call to examine the process that led to the widespread adoption of perioperative β blockade by many practitioners.10
A study of 10000 patients (POISE) is under way and plans to report early if a significant beneficial effect of β blockade is uncovered.11 More than 8000 patients have been recruited to the trial, which started in 2002 and is scheduled to finish in July 2008, but which may not achieve the target recruitment of 10000 patients. However, no results have been reported, suggesting that any beneficial effect of β blockers is likely to be moderate at best.11
Like β blockers, statins have also been advocated to reduce the risk of perioperative myocardial ischaemia. Despite studies involving nearly 800000 patients the number of people enrolled in randomised studies is small. The non-randomised studies suggest that statins confer benefit, but the evidence remains weak.5 The favourable results seen in cohort studies may be due to the beneficial effect of other agents taken concomitantly, rather than the effect of statins alone.
Randomised studies may prove valuable, but completing a multicentre randomised controlled trial like POISE will be challenging. To show that statins reduce the risk of myocardial events by 25%—which is a relatively low target, as the current literature suggests perioperative rates of death or acute coronary syndromes are 30-42% lower in statin users than in patients who are not taking statins at the time of surgery—a trial of at least 6000 people would be needed.5 For the same reduction in overall survival more than 12000 patients would be needed.5 12 The DECREASE IV trial plans to recruit over four years to assess the affects of a β blocker (bisoprolol) and a statin (fluvastatin), but it may face similar difficulties to those seen for the POISE trial.
The risks of myocardial events associated with sudden withdrawal of treatment are similar for β blockers and statins. However, while the safety profile of β blockers is well documented this is not so for statins, which are associated with serious liver and muscle toxicity, although these are rare in perioperative use.5 12
The benefits of statins in reducing myocardial ischaemic events in the general population and high risk patients are well known,5 12 but robust evidence to confirm that these drugs are valuable in routine perioperative use has not been published. So, on the basis of the evidence currently available what should practising clinicians do? We suggest that patients already receiving β blockers or statins before surgery should continue with treatment. Only patients who need heart rate or blood pressure control, or both, in the perioperative period should start treatment with β blockers. No patient should start taking statins in the perioperative period specifically to reduce the likelihood of perioperative cardiac events.
Competing interests: None declared.
Provenance and peer review: Commissioned; not externally peer reviewed.