Description of the condition
In non-cardiac surgery patients, postoperative myocardial ischaemia has been identified as a major predictor of cardiac complications (Mangano 1990b). During the intraoperative period, various factors could aggravate the burden on the cardiovascular system, such as the induction of the sympathetic nervous system by the surgery, use of anaesthetics, endotracheal intubation, and major blood loss. In addition, various anaesthetics cause vasodilation and depression of the cardiac muscle, which could result in decreased blood circulation and cardiac decompensation even in patients without cardiovascular disease. Furthermore, heart diseases (for example ischaemic heart disease, congestive heart failure (CHF)) are very common in patients undergoing preoperative anaesthesia check-ups. Cardiac complications are hard to avoid and are the main cause of perioperative morbidity and mortality in patients undergoing non-cardiac surgery. The most common and fatal cardiac complications during and after non-cardiac surgery include myocardial infarction (MI), pulmonary oedema, CHF, and arrhythmias.
Cardiac complications during non-cardiac surgery can lead to high morbidity and mortality, along with higher cost of patient care (Mangano 1990a; Mangano 1995). In 1991, Mangano et al first reported the incidence of intraoperative ischaemia (27%) and postoperative ischaemia (42%) in patients undergoing non-cardiac surgery in a clinical trial which included 100 patients with coronary artery disease or who were at risk of the disease (Mangano 1991). In another study, the authors reported that the incidence of cardiac events in patients who underwent non-cardiac surgery was 6.4%, which varied according to the difference in the American Society of Anesthesiologists (ASA) index stratum wherein the incidence of cardiac events in patients with ASA class-1 was 2.8% and increased to 50% in patients with ASA class- 5 (Gilbert 2000). The risk of perioperative MI in patients undergoing non-cardiac surgery is less than 1% (Fleisher 2001).
Cardiovascular diseases are the primary cause of morbidity and mortality in the US and Europe (Murray 1997; Yach 2004). Furthermore, with the increase in the aging population larger numbers of elderly patients are undergoing surgery, which predisposes them to higher cardiac risk. Thus more patients with or at risk of cardiac disease are undergoing non-cardiac surgery. In recent years, major American and European cardiac societies have issued guidelines for preoperative risk assessment and better patient care of non-cardiac surgery patients (ACC/AHA 2008; Hoeks 2010; JCS 2011). As a consequence, decreasing the cardiac risks and preventing associated complications are important issues among non-cardiac surgery patients.
Description of the intervention
Nitrates have been used in clinical practice for over 100 years (Brunton 1867; Murrell 1879). They are one of the most commonly used cardiovascular drugs. Organic nitrates are a large family and include a large number of drugs with a common molecular structure, -O-NO
How the intervention might work
The main pathophysiologic mechanism of nitrates is to produce nitric oxide (NO
- to provide relief from angina pectoris by improving the balance between myocardial oxygen consumption and oxygen supply;
- to abate heart failure by lowering cardiac preload and afterload;
- to treat cardiogenic pulmonary edema and pulmonary artery hypertension (PAH) by dilating pulmonary vessels and decreasing pulmonary vascular resistance (PVR).
Keeping all these factors in mind, prophylactic nitrates could play an important role in improving cardiac function in non-cardiac surgery patients by increasing myocardial oxygenation and furthermore reducing the risks of perioperative cardiac complications.
Why it is important to do this review
It is well-established that nitrates have definite efficacy for the management of MI, angina, and hypertension (Nossaman 2010). However, the efficacy of prophylactic nitrates in non-cardiac surgery is still undefined (ACC/AHA 2008; Kakisis 2003; Sear 2002).
There are only a few randomized controlled trials (RCTs) assessing the use of prophylactic nitrates during non-cardiac surgery to reduce cardiac risk (Stevens 2003), and they have not provided firm conclusions because of their small sample sizes.
We have not identified any systematic reviews or meta-analyses providing strong evidence on the prophylactic use of nitrates to prevent cardiac morbidity and mortality in patients undergoing non-cardiac surgery. Furthermore, there is no pooled evidence available for the appropriate dosage, time, and route of prescribed nitrates. Therefore, it is critical to review the evidence on the use of nitrates in patients undergoing non-cardiac surgery.
To assess the effects of nitrates as compared to other interventions or placebo for 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.
We also aim to identify the influence of different routes and dosages of nitrates on patient outcomes.
Criteria for considering studies for this review
Types of studies
We will include all parallel design and cluster RCTs. We define an RCT as a study in which patients are allocated to relevant groups according to a random method (as described in Higgins 2011), such as using a random numbers table, or randomized by statistical software. We will exclude non-randomized studies and those RCTs for which data are published only in abstract form and no further information can be obtained from the author.
Types of participants
We will include patients (aged 15 years and above) undergoing non-cardiac surgery under any type of anaesthesia.
We will exclude:
- patients with known heart disease (e.g. left bundle branch block) which can disturb observation of ST segment changes on the electrocardiogram (ECG);
- patients in whom nitrate use is unsuitable, such as those with hypotension, glaucoma, and obstructive myocardiopathy.
Types of interventions
We will include studies which compare nitrates as an intervention with:
- any other pharmacotherapy (e.g. beta-blockers, and calcium channel blockers); or
- placebo (defined as an inert substance designed to resemble the drug being tested but which has no active ingredient and has no treatment effect).
We will also include trials comparing different doses and routes (or both) of nitrate administration.
Types of outcome measures
We will include studies which report the effect of nitrates on cardiac risk and adverse events during non-cardiac surgery.
All-cause mortality at 30 days post-operation.
1. Perioperative incidence of cardiac morbidity, measured as incident episodes of any of:
- angina pectoris,
- acute myocardial infarction,
- acute heart failure,
- cardiac arrhythmia,
- cardiac arrest,
- increased troponin, and
- cardiac ischaemia indicated by ECG.
2. Any adverse event as a result of treatment, such as:
- side effects (e.g. gastrointestinal reaction, headache, and hypotension) from the start of observation to 24 hours post-operation,
- anaphylaxis, and
- treatment discontinuation caused by the intervention.
Search methods for identification of studies
We will search the following databases to identify all the published and unpublished trials meeting the inclusion criteria (see Criteria for considering studies for this review):
- Cochrane Central Register of Controlled Trials (CENTRAL) in The Cochrane Library (latest issue);
- MEDLINE (OvidSP) (1966 to date);
- EMBASE (OvidSP) (1980 to date); and
- Chinese Biomedical Base (CBM) (1982 to date).
We will use the search terms in Appendix 1 for MEDLINE and adapt them appropriately for the other databases, with no language restriction. We will arrange for potentially eligible non-English language papers to be translated so that they can be fully assessed for inclusion in the review.
Searching other resources
To identify any additional published, unpublished, and ongoing studies, we will:
- search conference proceedings of important anaesthesiology or cardiology scientific meetings, and abstracts;
- contact experts in the field for information on unpublished or incomplete research;
- handsearch speciality journals, such as journals published in China (e.g. Chinese Journal of Anesthesiology, the Journal of Clinical Anesthesiology,Chinese Journal of Cardiology, and Forum of Anesthesia and Monitoring);
- search Google Scholar for grey literature.
Data collection and analysis
Selection of studies
Two authors (NZ and TXW) will develop the search strategy to identify studies for inclusion in the systematic review. Two authors (NZ and JX) will independently assess the search results in phase one by reading the titles and the abstracts. We will resolve any disagreement by mutual consensus or with the input of a third author (TXW). Full articles of all the studies selected in phase one will be reviewed independently by two authors (NZ and JX). When needed, two authors (NZ and JX) will contact the study authors by telephone or email to obtain more information. We will scrutinize all the selected studies to check for multiple publications of the same trials. We will resolve any disagreement by mutual consensus or with input from a third author (TXW). Inter-reviewer agreement between the two authors during both the screening and study selection phases will be assessed by Cohen's kappa statistic (Cohen 1960).
Data extraction and management
Two independent authors (NZ and JX) will extract data from the included studies to minimize errors and reduce potential biases. We will use a pre-designed electronic data abstraction form (Appendix 2) for the data extraction. We will extract information about the study source, eligibility, methods, participants, interventions, outcome measures, and other relevant information as reported. We will try to extract detailed data as much as possible by searching the full texts or interviewing the original author directly, or both. If a study is reported in more than one publication, we will extract data from each publication separately, then combine the information across multiple datasets. We will resolve any discrepancy during the data extraction process by mutual discussion among all the authors.
Assessment of risk of bias in included studies
Two authors (NZ and JX) will independently assess the risk of bias of the included trials using the Cochrane 'Risk of bias' tool (Higgins 2011). We will assess the following domains:
- sequence generation (selection bias),
- allocation concealment (selection bias),
- blinding of participants and personnel (performance bias),
- blinding of outcome assessment (detection bias),
- incomplete outcome data (attrition bias), and
- selective outcome reporting (reporting bias).
We will assess the domains as ‘low risk’, ‘high risk', or ‘unclear risk’. We will resolve any disagreements by mutual consensus. We will report the results of our assessment by creating a ‘Risk of bias’ graph and a ‘Risk of bias’ summary figure using Review Manager (RevMan 5.1) software, if appropriate. We will also report the risk of bias in the 'Results' section. We will provide summary assessments of the risk of bias for each outcome within and across studies. We will assess inter-reviewer agreement between the two authors by Cohen's kappa statistic (Cohen 1960).
Measures of treatment effect
For dichotomous outcomes, we will express the results as risk ratios (RR) with 95% confidence intervals (CI). For continuous data, we will report the standardized mean difference (SMD).
Unit of analysis issues
We will include studies which report the effect of nitrates on cardiac risk and adverse events during non-cardiac surgery, therefore, both the unit of analysis and the level of randomization should be individual. If we include cluster randomized trials, and the level of randomization is not by individual but group, we will try to obtain the relevant patient-level data by contacting the study author. We will then extract and analyse the data at the level of the individual. If we are unable to obtain the patient-level data, we will conduct a subgroup analysis according to the different data unit.
Dealing with missing data
In the case of missing data, we will try to contact the trial author to request the missing data. If this is unsuccessful, we will apply appropriate statistical methods to deal with the omission. If there are missing standard deviations we will try to calculate or estimate them using imputation techniques (for example using confidence intervals, standard errors, t values, P values, F values). In addition, we will conduct sensitivity analyses to observe the influence of the missing data. Finally, in the 'Discussion' section, we will explore the potential impact of missing data on the findings of the review.
If participants have been excluded from the analysis, such as through loss to follow-up, we will conduct an intention-to-treat (ITT) analysis.
Assessment of heterogeneity
We define heterogeneity as including all the kinds of variability among studies in a systematic review. It can be categorized as clinical heterogeneity, methodological heterogeneity, and statistical heterogeneity (Higgins 2011). The statistical heterogeneity among the included studies will be quantified with the I
In cases of clinical heterogeneity, we will determine whether meta-analysis is appropriate through discussion among all the review authors, or by asking other clinical experts.
Assessment of reporting biases
Firstly, to minimize the impact of reporting biases as far as possible, we will ensure a comprehensive search for eligible studies by conducting our strategy listed as above. We will examine funnel plots to assess the potential for publication bias if we have 10 or more studies reporting on a particular outcome (Egger 1997). If there is asymmetrical appearance of the funnel plot, we will suspect publication bias after excluding other causes for the asymmetry. We will try to explore the cause of publication bias in the discussion.
We will synthesize data using Review Manager version 5.1 (RevMan 5.1). The model used to pool the data depends on the existence and extent of heterogeneity. When there is no statistical heterogeneity or low heterogeneity (I
If the number of included studies is too small (for example we only find one RCT) or heterogeneity is apparent, meta-analysis is neither appropriate nor possible. Under those circumstances we will narratively report the results of our systematic review.
If included studies have multiple arms, we will identify the relevant intervention and control groups and combine the relevant intervention groups into a single group, and the relevant control groups into a single group, before synthesizing the data (Higgins 2011).
Subgroup analysis and investigation of heterogeneity
If heterogeneity is obvious among our studies, we will explore possible sources of heterogeneity and judge whether to conduct subgroup analyses. These will focus on two aspects:
- the presence of baseline cardiac disease before the surgery, because nitrates are commonly used to treat Ischaemic heart disease (IHD) so the percentage of IHD patients in the study may influence the results;
- the type of surgery (high risk or major vascular surgery compared to low risk or minor abdominal surgery can lead to rapid haemodynamic changes).
We will undertake sensitivity analysis to identify the robustness of results, considering the following factors.
- Methodological design: allocation concealment or blinding (adequate, unclear, inadequate, or not performed).
- Patient characteristics: age, IHD, major surgery, other treatment measures except the experimental intervention.
- Intervention: dose, route of administration, control group (placebo or active intervention).
- Follow-up outcome (loss to follow-up > 20%).
Summary of findings
We will assess the quality of the body of evidence associated with specific outcomes (perioperative death, acute MI, acute heart failure, and adverse events) in our review according to the principles of the GRADE system (Guyatt 2008), and we will construct a 'Summary of findings' (SoF) table using the GRADE software.
We thank Jane Cracknell (managing editor), Arash Afshari (content editor), Cathal Walsh (statistical editor), Janet Wale (copy editor), Pierre Foex, Javier Eslava-Schmalbach, and Tina Hu (peer reviewers) for their help and editorial advice during the preparation of this protocol for the systematic review.
Appendix 1. Ovid MEDLINE search strategy
1. (non?cardiac adj3 (surg* or operat*)).af. or perioperative.mp. or exp Perioperative Period/ or Postoperative Complications/ or complication*.ti,ab. or (cardiac adj3 (complication* or mortality or morbidity or risk)).mp.
2. exp Nitrates/ or exp Nitroglycerin/ or (Nitroglycerin* or NTG or (glyceryl adj3 trinitrate*) or Trinitrin or Anginine or Isosorbide dinitrate or Isosorbide mononitrate or (isosorbide adj3 (dinitrate or mononitrate)) or Isoket or GTN).mp.
3. 1 and 2
4. ((randomized controlled trial or controlled clinical trial).pt. or randomized.ab. or placebo.ab. or clinical trials as topic.sh. or randomly.ab. or trial.ti.) not (animals not (humans and animals)).sh.
5. 3 and 4
Appendix 2. Data extraction form
1. Review information
2. General information
More than one report of the same study
3. Participant and trial characteristics
* If cross-over design, please refer to the Cochrane Editorial Office for further advice on how to analyse these data
5. Risk of biases
Contributions of authors
Conceiving the review: Na Zhao (NZ), Jin Xu (JX)
Co-ordinating the review: NZ, Balwinder Singh (BS)
Undertaking manual searches: NZ, JX
Screening search results: NZ, JX
Organizing retrieval of papers: NZ
Screening retrieved papers against inclusion criteria: NZ, JX, Taixiang Wu (TXW)
Appraising quality of papers: NZ, JX, TXW
Abstracting data from papers: NZ, JX
Writing to authors of papers for additional information: NZ
Providing additional data about papers: NZ
Obtaining and screening data on unpublished studies: NZ, JX
Data management for the review: NZ
Entering data into Review Manager (RevMan 5.1): NZ, JX
RevMan statistical data: NZ, JX, TXW
Other statistical analysis not using RevMan: NZ, TXW
Interpretation of data: NZ, JX, TXW, Xuerong Yu (XRY), BS
Statistical inferences: NZ, JX, TXW, BS
Writing the review: NZ, JX, TXW, XRY, BS
Guarantor for the review (one author): NZ
Person responsible for reading and checking review before submission: JX, XRY, BS
Declarations of interest
Na Zhao: none known
Jin Xu: none known
Balwinder Singh: none known
Taixiang Wu: none known
Xuerong Yu: none known