In this superiority, randomized, controlled, single-blinded, single-center trial, we plan to include patients scheduled to undergo elective major surgery (plastic, visceral, urologic, gynecologic or ENT surgeries). The patients will be allocated to receive either an OFA protocol or a standard anesthesia protocol (Fig. 1). The study will be conducted at the Angers University Hospital, Angers, France. This trial is designed following the Standard Protocol Items (SPIRIT guidelines). Figures 2 and 3 provide an overview of the study plan.
The French Institutional Review Board “Eastern III” (Nancy, France) and the National Agency for Drug and Health Product Safety have approved the study protocol (2021-A00364-37) in its version 1. Patients will provide written consent for participation. The study will be conducted in accordance with the Declaration of Helsinki.
Participants
Eligible patients are adults (age ≥ 18 years) undergoing a scheduled surgery with an estimated time of 90 minutes or more and supposed to be painful (with the usual need for postoperative intravenous (IV) morphine patient-controlled analgesia (PCA)). The surgical specialties concerned are ENT, plastic, digestive, urologic and gynecologic surgery. Eligible patients must be French-speaking and must have sufficient cognitive ability to complete a questionnaire. The exclusion criteria are surgery with an orthopedic procedure (e.g. osteosynthesis), need for rapid sequence anesthetic induction, pregnancy or lactation, severe psychiatric or cognitive disorder interfering with questionnaire assessment, body mass index < 18 or > 39 kg/m2, any contraindication to study drugs, porphyria, uncontrolled epilepsy, unstable cardiac insufficiency, chronic renal or hepatic failure, preoperative bradycardia with an atrioventricular block (2nd or 3rd degree), chronic beta-blocker treatment, ongoing treatment with a QT interval prolonging drug.
Information of patients
During the anesthesia consultation, investigators will verify inclusion/exclusion criteria. The investigator will invite the patients to participate. Patients will receive complete information in faithful terms and understandable language concerning the objectives, the required follow-up, the risks, the safety measures and the right to refuse to participate or stop the study at any time. The investigator obtained written informed consent signed by both the investigator and the patient.
Randomization and blinding
The included patients will be randomized to one of the two groups (OFA or standard of care) using a web-based system (Ennov Clinical® software) by the investigator anesthesiologist in charge on the day or the day before the scheduled surgery.
Patients will be randomized in one of the two groups with a 1:1 ratio, using a minimization algorithm defined by the study methodologist, based on two factors: the type of surgery (ENT, plastic, digestive, urologic and gynecologic surgery) and the surgery severity (assessed by the Surgical Outcome Risk Tool (SORT) score, which can be classified in minor, intermediate, major and major/complex surgery, and which is available online at www.sortsurgery.com).
The anesthesia team performing the general anesthesia will be unblinded on the allocation group. The patient will be kept blinded to the protocol allocation, as the nurses and the medical team involved in the postoperative care and outcomes evaluation.
Intervention
The study aims to compare an OFA protocol to an anesthesia protocol based on standard practices. In both arms, patients will undergo general anesthesia, combined or not with local or regional anesthesia. The two protocols are detailed in Fig. 1.
None of the patients will receive premedication. Anesthetic monitoring will be the same in both groups, including pulse oximetry, electrocardiography, non-invasive blood pressure and/or invasive monitoring when indicated, body temperature, muscle relaxation (train-of-four stimulation), and bispectral index.
The OFA protocol (OFA group) begins before surgery with a clonidine infusion (under cardiac monitoring) at an initial rate of 50 µg/h, then adapted according to hemodynamic stability (to keep systolic blood pressure in a range of ± 20% its basal value) with a maximal rate of 150 µg/h (100 µg/h for patients with body weight < 50 kg). Anesthesia will be induced and maintained with hypnotic drugs and curare left at the anesthesiologist's discretion. No opioids should be infused. The clonidine infusion should be continued at the same infusion rate. The OFA protocol includes a magnesium sulfate infusion (40 mg/kg), a lidocaine infusion (1.5 mg/kg bolus dose in ten minutes followed by 1.5 mg/kg/h continuous infusion), and a ketamine infusion (0.5 mg/kg bolus dose followed by a 0.2 mg/kg/h continuous infusion). The ketamine infusion should be stopped thirty minutes before the end of the surgery, and the lidocaine infusion will be stopped after one hour in the postanesthesia care unit (PACU).
The standard anesthesia protocol associates hypnotic drugs, curare and opioids (sufentanil or remifentanil) left at the discretion of the anesthetist in charge. A low-dose ketamine bolus (0.15 mg/kg) is allowed at anesthesia induction, possibly followed by repeated boluses since it is commonly used in our standard practice.
In both arms, according to recent recommendations, anesthesiologists are prompted to infuse an antimicrobial prophylaxis [17], to apply a protective ventilation strategy (tidal volume between 6–8 ml/kg, respiratory rate for an end-tidal CO2 value at 35–40 mmHg, a positive end expiratory pressure) [18], and to treat perioperative hypotension and hypovolemia [19]. The depth of anesthesia is monitored by a bispectral index sensor (target: 40–60). Postoperative pain, nausea and vomiting are managed according to the unit protocols. The use of local or regional anesthesia is allowed in both groups (in the OFA group, the protocol stipulates not to use opioids as adjuvant). In the PACU, if the patient experiences acute pain (defined as a pain analog scale ≥ 4), morphine or oxycodone titration will be started (bolus of 1 to 2 mg, with a 5-minute delay between two doses) until pain relief is achieved. The analgesic management will be continued in the ward using IV PCA and/or oral opioids. The doses of opioids (i.e., morphine and oxycodone) will be converted into morphine-equivalent doses according to the following rule: 1 mg oral morphine = 1/2 mg oral oxycodone = 1/3 mg IV morphine; 1 mg IV morphine = 1 mg IV oxycodone. Patients will be discharged to the ward when their modified Aldrete score is ≥ 12 [20].
Outcomes
Primary outcome
The primary outcome is the early postoperative quality of recovery, assessed via the French version of the QoR-15 questionnaire score (FQoR-15) [21], and measured at 24 hours after the surgery (postoperative day 1, POD 1). The QoR-15 score is currently the most reliable and reproducible tool to determine the quality of postoperative recovery [22–24]. Its use is recommended to assess the patient’s well-being, according to a recent international consensus [25], and has already been assessed by previous studies [26].
The FQoR-15 score is obtained via a self-administered questionnaire (which can also be performed through a phone interview) and consists in 15 items scored on an 11-point scale, with a total score (sum of each item) range from 0 to 150 (0 for very bad recovery and 150 for an excellent quality of recovery). The FQoR-15 questionnaire is given by the nurses in the ward on a paper sheet and is filled by the patient.
Secondary outcomes
In order to confirm the impact of the OFA protocol on postoperative recovery, we will also measure the QoR-15 score at 48 hours and 72 hours after the surgery in all patients. We will evaluate the pain during effort (mobilization, cough or physiotherapy sessions) and the number of postoperative complications according to the POMS (PostOperative Morbidity Survey) classification [27], assessed up to 72 hours after the surgery. The postoperative consumption of morphine-equivalent will also be compared between the two groups. The lengths of stay in PACU and in the hospital will be measured.
We will compare the patient tolerance to the allocated protocol with the assessment of the hemodynamic status (intra-operative use of vasopressor or cessation of one of the drugs for hemodynamic purposes), heart-rhythm (< 35 or > 140 beats per minute for more than 30 seconds) and anaphylactic reactions. Another secondary outcome is the respect of the allocated protocol by the anesthesia team: OFA protocol will be considered complete if at least two of the following drugs are used between ketamine, lidocaine, clonidine and magnesium sulfate, and if no intraoperative opioids are used; the standard group will be considered complete if lidocaine, clonidine or magnesium sulfate are not used intra-operatively. The surgeon’s and anesthesiologist’s satisfactions concerning the anesthesia will be assessed after the surgery according to a numerical scale (rated from 0 to 10). We will also compare the patient satisfaction of anesthesia at 24 hours via a numerated scale from 0 to 10 (0 = not satisfied and 10 = totally satisfied) and the presence or absence of intra-operative memorization.
Three months after surgery, a phone interview will be conducted to evaluate the chronic pain using the Brief Pain Inventory (BPI) [28] in its French short version, the quality of life using the French EQ-5D-3L questionnaire (available on request on https://euroqol.org/), and the rate of neuropathic pain diagnosed using the DN-4 questionnaire [29].
Sample size
According to the database used to assess the French version of the QoR-15, that included 72 patients undergoing gynecologic, digestive or urologic major surgery, the mean QoR-15 score was 97 at 24 hours after the surgery, with a standard deviation of 16. A difference of at least 8.0 points in the score is considered clinically significant [30]. The OFA will thus be superior to the standard of care if the mean QoR-15 is at least 105. Considering an alpha risk of 0.05 and a power of 0.8, the estimated number of patients needed is 126 (63 patients in both groups). Considering that data will not be available for ten percent of the patients (e.g., surgery cancelled, lost in follow-up, consent withdrawal) regarding the primary outcome, we plan to include 140 patients.
Follow-up
Nine visits are scheduled for all patients enrolled in the study. The plan summarizing the follow-up visits is presented in Fig. 2. The time T0 (hour 0/day 0) is the time of the surgical incision. The following data will be recorded at inclusion: demographic data, weight and height, ASA physical status classification, surgery type, comorbidities. The SORT classification will also be recorded.
Perioperative Surgical and anesthetic data will be recorded, including total dose of each medication, tolerance and safety data (i.e., hypotension, bradycardia, tachycardia, need of vasopressor drugs, oxygen desaturation (i.e., SpO2 < 92% for at least 5 minutes), or hypersensitivity reactions). The anesthesiologist and the surgeon will also grade their satisfaction concerning the anesthesia.
In PACU, the total dose of pain medications, the occurrence of any complication, including postoperative nausea and vomiting (PONV), numerical pain rating scale, and PACU length of stay (in hours) will be recorded. In the ward visits (at 24, 48 and 72 hours after the surgery), we will record the following data: QoR-15 questionnaire, maximal numerical effort pain scale, the occurrence of complications (using the POMS classification), morphine consumption, bowel movement, adverse event and patient satisfaction (only at 24 hours after the surgery).
At seven days, we will collect the morphine or equivalent daily consumption and the occurrence of surgical revision. The following data will be recorded at the hospital discharged visit (or 30 days after surgery): total length of stay and occurrence of in-hospital opioid side effects. A phone call to the patient will be made at 3-month to record the BPI, the DN-4 and the EQ-5D scores.
Data collection and study monitoring
Data will be entered into the electronic web-based case report form (eCRF on Ennov Clinical®). We will establish the database from the eCRFs.
A Clinical Research Associate (CRA) mandated by the study sponsor will ensure the successful completion of the study, the collection of data, documentation, recording and report, in accordance with the Standard Operating Procedures implemented in the Angers Hospital and in accordance with Good Clinical Practice, laws and regulations.
The following items will be reviewed:
-
Signed informed consent,
-
Compliance with the study protocol and procedures that are defined,
-
Quality of data collected in the case report form: accuracy, missing data, data consistency with the documents ‘source’ (medical records, appointment books, original lab results, etc.).
After monitoring data on site, an automated check of the data entered will be made by the data management team based on the data validation plan signed by the coordinating investigator. Detected errors will lead to the issuance of requests for information and electronic correction.
Statistical analysis
All analyses will be performed with R software (version 3.6.3). The main analysis will be in intention-to-treat and will include all randomized patients. Patients will be analyzed according to their randomization group. A per-protocol analysis will be performed as a sensibility analysis to assess the main analysis robustness. A flow chart of all patients (Fig. 3) and descriptive statistics will be used to describe baseline characteristics. Data will be presented by their mean with standard deviation, median with interquartile range, number with the percentage of sample (%), or 95%-confidence interval, according to their normal or non-normal distribution.
For the primary endpoint, we will use multivariate imputation for chained equations (with five imputations) via the mice package in R, and perform a pooled-linear regression model, allowing us to perform an intention-to-treat analysis even if missing data. The covariates included in this analysis will be the allocation group, the type and the surgery severity. This statistical method will also be used to analyze the QoR-15 at 48 and 72 hours as secondary endpoints. We will also impute data for the analyses of other secondary endpoints: the rate of chronic pain patients at three months identified by the BPI via a logistic regression (same adjustment covariates) as well as for EQ-5D visual analog score via a linear regression (same adjustment covariates). The other endpoints will not be analyzed with the multiple imputation method. We reserve imputation for the endpoints included in the hierarchical scheme of alpha risk management.
Concerning the pain evolution and morphine consumption variation, we will analyze those outcomes using a mixed regression model (fixed effect for the treatment arm, a fixed effect for the time frame, a fixed effect for each minimization factor, and a random effect for the patients). Hospital and PACU length of stays, as well as satisfactions (those of the surgeon, those of the anesthesiologist, those of the patient), will be compared using a linear regression model (allocation group and minimization factors as covariates). The proportion of complications of the POMS will be defined as a categorical variable (four levels: no complication, one complication, two complications, and more than 3 complications) and will be analyzed using a logistic regression model (allocation group and the minimization factors as covariates). The rate of patients presenting neuropathic pain, as well as the proportion of patients experiencing memorizations, will be compared by logistical regression models (allocation group and the minimization factors as covariates). Pain characteristics described by BPI and the proportion of patients in each of the five scales of EQ-5D will not be compared but only described.
All statistical analyses will be performed by two-sided tests. The signification cut-off will be a p-value < 0,05 with 95% confidence interval.
Alpha risk management
We will manage the alpha risk in a sequential hierarchical manner (Fig. 4). If the p-value is less than 0.05 for the assessment of the QoR-15 at H24, we will continue the analysis at H48, then at H72 (if the p-value remains less than 0.05 at H48). We will continue to assess our alternative hypothesis on the proportion of chronic pain patients (if the p-value remains less than 0. 05 for the QoR-15 at H72), and finally we will look at the quality of life by the EQ-5D visual analog scale (if the p-value remains below 0.05 for the proportion of chronic pain patients). If at one point, the p-value is > 0.05, the comparisons will only be made for exploratory purposes and will be delivered explicitly as such. Comparisons made on the other endpoints will be for exploratory purposes.