Na Zhao, Jin Xu, Balwinder Singh, Xuerong Yu, Taixiang Wu, Yuguang Huang

Cardiac complications are not uncommon in patients undergoing non-cardiac surgery, especially in patients with coronary artery disease (CAD) or at high risk of CAD. Perioperative cardiac complications can lead to mortality and morbidity, as well as higher costs for patient care. Nitrates, which are among the most commonly used cardiovascular drugs, perform the function of decreasing cardiac preload while improving cardiac blood perfusion. Sometimes, nitrates are administered to patients undergoing non-cardiac surgery to reduce the incidence of cardiac complications, especially for patients with CAD. However, their effects on patients' relevant outcomes remain controversial. • To assess effects of nitrates as compared with other interventions or placebo in reducing cardiac risk (such as death caused by cardiac factors, angina pectoris, acute myocardial infarction, acute heart failure and cardiac arrhythmia) in patients undergoing non-cardiac surgery.• To identify the influence of different routes and dosages of nitrates on patient outcomes. We searched the Cochrane Central Register of Controlled Trials (CENTRAL), MEDLINE, EMBASE and the Chinese BioMedical Database until June 2014. We also searched relevant conference abstracts of important anaesthesiology or cardiology scientific meetings, the database of ongoing trials and Google Scholar.We reran the search in January 2016. We added three potential new studies of interest to the list of 'Studies awaiting classification' and will incorporate them into our formal review findings for the review update. We included randomized controlled trials (RCTs) comparing nitrates versus no treatment, placebo or other pharmacological interventions in participants (15 years of age and older) undergoing non-cardiac surgery under any type of anaesthesia. We used standard methodological procedures as expected by Cochrane. Two review authors selected trials, extracted data from included studies and assessed risk of bias. We resolved differences by discussion and, when necessary, sought help and suggestions from a third review author. We used a random-effects model for data analysis. We included 27 randomized controlled trials (RCTs) (8244 participants analysed). Investigators reported 12 different comparisons of three different nitrates (nitroglycerin, isosorbide dinitrate and nicorandil) versus no treatment, placebo or other pharmacological interventions. All participants were older than 15 years of age. More than half of the trials used general anaesthesia. Surgical procedures in most trials were at low to moderate risk for perioperative cardiac complications. Only two comparisons including three studies reported the primary outcome - all-cause mortality up to 30 days post operation. Researchers reported other morbidity outcomes and adverse events in a variable and heterogeneous way, resulting in limited available data for inclusion in the meta-analysis. We determined that the overall methodological quality of included studies was fair to low, in accordance with risk of bias in most domains.In summary, we found no difference in the primary outcome - all-cause mortality up to 30 days post operation - when nitroglycerin was compared with no treatment (one study, 60 participants, 0/30 vs 1/30; (risk ratio (RR) 0.33, 95% confidence interval (CI) 0.01 to 7.87, very low-quality evidence based on GRADE criteria) or with placebo (two studies, 89 participants, 1/45 vs 0/44; RR 2.81, 95% CI 0.12 to 63.83, very low-quality evidence). Regarding our secondary outcomes, we noted no statistically significant differences in angina pectoris, acute myocardial infarction, acute heart failure, cardiac arrhythmia or cardiac arrest in any comparisons. In comparisons versus nitroglycerin, although more events of cardiac ischaemia were observed in participants receiving no treatment or placebo, we found no statistically significant differences in any comparisons, except the comparison of nicorandil versus placebo. One study revealed a potential dose-dependent protective effect of nicorandil for cardiac ischaemia.Adverse events were reported in a heterogeneous way among the comparisons. In general, more participants treated with nitrates had hypotension, tachycardia and headache, but investigators reported no statistically significant differences between groups in any comparisons. This systematic review suggests that nitroglycerin or isosorbide dinitrate is not associated with improvement in mortality and cardiac complications among patients undergoing non-cardiac surgery. Limited evidence suggests that nicorandil may reduce the risk of cardiac ischaemia in participants undergoing non-cardiac surgery. Additional studies are needed to consolidate the evidence.However, the data included in many of the analyses in this review are sparse - that is, adequate data are few - resulting in very low power to detect differences between nitrates and comparators. Thus, a more objective conclusion would state that available evidence is insufficient to show whether nitrates are associated with improvement in mortality and cardiac complications among patients undergoing non-cardiac surgery.Over the past decade, no high-quality studies have focused on association of cardiac mortality and morbidity with use of nitrates during non-cardiac surgery. This review underlines the need for well-designed trials in this field.