Trial Design
Across two centres (both a large quaternary public hospital and private hospital setting), adults admitted for surgical operations requiring muscle paralysis were enrolled in a prospective, double-blind, randomised controlled trial comparing specific outcomes following reversal with sugammadex or neostigmine. The study was approved by the Northern Sydney Local Health District Human Research Ethics Committee (RESP/16/289) and the North Shore Private Hospital Ethics Committee (2017-001).
The trial was registered at the Australian New Zealand Clinical Trials Registry (www.anzctr.org.au) # ACTRN12616000063415 (21/01/2016) and The U.S. National Library of Medicine (www.ClinicalTrials.gov) #NCT02825576 (07/07/2016). The trial was compliant with the Australian National Health and Medical Research Council (NHMRC) statement on ethical conduct in human research (2007) (14) and the note for guidance on good clinical practice (CPMP/ICH-135/95) (15).
Participants
Screening for eligible patients was conducted by examining operating theatre and preadmission clinic schedules for each day. Inclusion criteria were: patients over 18 years of age, planned to undergo non-cardiac surgery with an expected operative time of >2 hours, use of endotracheal intubation, and an expected hospital length of stay greater than one night. These criteria were designed to preferentially select patients who would likely be at intermediate or high risk for PPCs (16).
Exclusion criteria were: patients previously recruited, patient weight >200kg, planned expectation to remain intubated postoperatively, and those with known hypersensitivity reactions or contraindications to any of the study drugs.
Patient recruitment occurred during a preoperative consultation, either in a preadmission clinic or on a hospital ward, prior to their transfer to the operating theatre.
Interventions
Patients were randomly assigned into one of two groups; the sugammadex group or the neostigmine group. At the end of surgery, the participant received a blinded reversal dose via a 10ml syringe with administration of 1mL/10kg of the blinded solution. The solution was standardised to contain the protocol’s reversal dose that would equate to either 2mg/kg of sugammadex or 50mcg/kg neostigmine with 10mcg/kg glycopyrrolate.
The study protocol required the use of intraoperative neuromuscular twitch (NMT) monitoring to ensure return of a train-of-four count (TOFC) greater than or equal to 2 prior to reversal, and restricted the use of neuromuscular blockers to rocuronium or vecuronium, at the choice of the individual anaesthetist.
There were no restrictions on the mode of anaesthesia, analgesia, use of postoperative nausea and vomiting (PONV) prophylaxis, or timing of neuromuscular blockade reversal.
Randomisation
Patients were allocated to treatment group via a computer-based simple randomisation software (randomizer.org). Group allocations were placed in a sealed opaque envelope by a research team member who was not involved in patient treatment or follow up. Prior to reversal, the treating anaesthetist recruited a member of the hospital clinical staff to open the next available randomisation envelope, draw up the allocated reversal agent based on the enclosed instructions, and deliver this to the treating anaesthetist. This individual was then not involved in further management of the trial patient.
Blinding
The patient, treating anaesthetist, all care providers, and research personnel assessing patient outcomes were all blinded to the patient’s treatment allocation.
Outcomes
The primary outcome was the incidence of PPCs (as defined by the European Perioperative Clinical Outcome (EPCO) statement (17) in the first three days postoperatively (Table 1).
Secondary outcomes were; patient reported quality of recovery at day 1 and 30 utilising the QoR-15 (18), the incidence of PONV, and the incidence of a defined set of airway events in the post anaesthetic care unit (PACU). In addition to the Quality-of-Recovery-15 score (QoR-15) at day 30, patients were also asked if they had also received either antibiotics or bronchodilators to treat a chest infection (or if previously on bronchodilators, whether they had increased their usage or dose) in the interim. The hospital discharge summary was reviewed to identify the diagnosis of a respiratory infection or associated complications and to calculate hospital length of stay.
Subgroup analysis was constructed to facilitate stratification into groups based on the risk of PPC, as defined by the Assess Respiratory Risk in Surgical Patients in Catalonia (ARISCAT) score (16), the risk of PONV as defined by the Apfel score (19), the use of Total Intravenous Anaesthesia (TIVA) vs volatile based anaesthesia, patient Body Mass Index (BMI), and American Society of Anesthesiologists (ASA) Class.
Table 1 outlines the EPCO definition, ARISCAT and Apfel scores, PACU Airway Events and the PONV scoring systems used.
Table 1 – Definitions used for outcome measurements and risk stratification
European Perioperative Clinical Outcome (EPCO) definitions of Post-operative Pulmonary Complications (17)
Respiratory infection
|
Patient has received antibiotics for suspected respiratory infection and met one or more of the following criteria: new or changed sputum, new or changed lung opacities, fever, white blood cell count >12×109/litre
|
Respiratory failure
|
Postoperative PaO2 <60 mm Hg on room air, PaO2:FiO2 ratio <300 mm Hg), or arterial oxyhaemoglobin saturation measured with pulse oximetry <90% and needing oxygen therapy
|
Pleural effusion
|
Chest radiograph showing blunting of costophrenic angle, loss of sharp silhouette of ipsilateral hemidiaphragm in upright position, evidence of displacement of adjacent anatomical structures, or (in supine position) hazy opacity in one hemithorax with preserved vascular shadows
|
Atelectasis
|
Lung opacification with shift of mediastinum, hilum, or hemidiaphragm towards affected area, and compensatory over-inflation in adjacent non-atelectatic lung
|
Pneumothorax
|
Air in pleural space with no vascular bed surrounding visceral pleura
|
|
Assess Respiratory Risk in Surgical Patients in Catalonia (ARISCAT) Score for risk prediction of postoperative pulmonary complications (16)
Risk Factor
|
Score
|
Age 51-80
|
3
|
Age >80
|
16
|
Preoperative SpO2 91-95%
|
8
|
Preoperative SpO2 ≤ 90%
|
24
|
Respiratory infection in past 1 month
|
17
|
Preoperative Haemoglobin<10 gm/dl
|
11
|
Upper abdominal incision
|
15
|
Intrathoracic incision
|
24
|
Surgery Duration 2-3 hours
|
16
|
Surgery Duration >3 hours
|
23
|
Emergency Procedure
|
8
|
Total Score - Low risk: <26, Intermediate risk: 26-44, High risk: ≥45
|
Apfel Score for risk prediction of PONV (19)
Risk Factors (1 point each) – Female sex, History of PONV or motion sickness, non-smoker, postoperative opioid treatment planned.
Total score and risk stratification
0
|
Minimal Risk
|
1
|
Low Risk
|
2
|
Intermediate Risk
|
3
|
High Risk
|
4
|
Very High Risk
|
|
Definitions of postoperative care unit events
PACU Events
- Any desaturation to SpO2<90%
- Need for manual airway support
- Need for oropharyngeal or nasopharyngeal airway
- Need for reintubation in PACU
- Need for anaesthetist to review the patient
- Unplanned ICU admission
PONV score
- no PONV
- PONV responsive to antiemetics
- PONV unresponsive to antiemetics
|
PaO2 -partial pressure of arterial oxygen, FiO2 – Fraction of inspired oxygen, SpO2 – peripheral capillary oxygen saturation, PONV – Postoperative Nausea and Vomiting, PACU – Post Anaesthesia Care Unit, ICU – Intensive Care Unit
Sample Size
Prior to the design of this trial, the largest published prospective study of PPCs showed an overall incidence of 7.9%, stratified into low risk (3.4%), intermediate risk (13.0%), and high risk (38.0%) (16).
Given this study was designed to include intermediate and high-risk patients, a conservative estimate of a 7% baseline incidence of PPCs was used. Based on prior retrospective work, showing an odds ratio of 0.28 (20), sugammadex would be expected to reduce the incidence of PPCs from 7–1.96%. Choosing a conservative reduction to an incidence of 3.5%, produced a clinically relevant Number Needed to Treat (NNT) of 29. Accepting an alpha error of 0.05, a beta error of 0.2, and estimating 5% incomplete data and loss to follow up resulted in a target sample size of 976 patients (21).
Statistical analysis
The presence of a PPCs, and the presence of PACU events were assessed as categorical variables, QoR-15 score and hospital stay were assessed as continuous variables, and PONV score was assessed as an ordinal variable. Continuous variables were assessed with two tailed t-tests, categorical and ordinal data was assessed by Chi-squared analysis of contingency tables (or Fisher’s exact test when n<5 in any outcome).
Binomial regression analysis was planned for categorical outcomes in subgroup analyses, with logistic regression for the ordinal variables.
Interim analysis on the primary outcome was planned when 488 (50% of patients) had been randomised and had completed the 30 day follow-up. The Haybittle-Peto approach was planned to be used, where the trial will be ended with symmetric stopping boundaries at P < 0.001. The trial was overseen by a Data and Safety Monitoring Committee (DSMC) who would advise on early trial cessation in case of futility of treatment effect or safety concerns.