Study design
The IMPAKT ERAS trial was approved by Investigational Review Board (IRB) at Vanderbilt University Medical Center (VUMC), and all methods were carried out in accordance with relevant guidelines and regulations. The trial will be a single-center, pragmatic, double-blind, cluster-randomized, placebo-controlled study. Adult patients presenting for an elective abdominal surgery through the colorectal, surgical oncology, or large ventral hernia ERAS services at a major academic institution will be eligible for enrollment. Major deviations from the standard ERAS protocols will be excluded, including the inability/patient refusal of a regional or neuraxial nerve block, a direct admission of an intubated patient to the Intensive Care Unit (ICU) from the operating room, and abortion of the planned surgical procedure (Box 1). Patients will be randomized to receive either a blinded ketamine or saline placebo administered intraoperatively as a bolus and infusion, followed by a postoperative infusion for up to 48 hours. All other elements of the ERAS pathways will remain the same between the two groups.
Box 1: Inclusion and exclusion criteria for the IMPAKT ERAS trial
Inclusion criteria
|
Adult patients undergoing an elective major abdominal surgery within a colorectal, surgical oncology, or ventral hernia enhanced recovery pathway.
|
Exclusion criteria
|
• Participant declines to consent for participation
• Contraindications to, or patient refusal, of standard enhanced recovery elements (e.g. nerve block)
• Surgery performed on a weekend day (non-elective)
• Direct postoperative admission from the operating room to the Intensive Care Unit with an endotracheal tube in place
• Cancellation or abortion of the planned surgical procedure
• Perioperative clinician deems the patient not appropriate for inclusion
|
Pragmatic features
The IMPAKT ERAS trial is supported by the Learning Healthcare System (LHS) at VUMC, which provides a supportive platform to integrate clinical practice with pragmatic research principles. The IMPAKT ERAS trial was intentionally designed to function within the constraints of routine clinical care at our institution where perioperative ketamine infusions have been utilized as a core component of our ERAS bundles since 2016. Therefore, the study protocols will be performed by clinical anesthesia providers (faculty, nurse anesthetists, and anesthesiology residents) without a need for additional training or team of dedicated research staff. The IRB deemed the study to be of minimal risk and approved a shortened, modified research consent process that is included within the routine clinical analgesia consent obtained by the Perioperative Medicine Team. Subjects provided written consent to participate. Study outcomes will be collected through the electronic medical record (EMR). (PRECIS Table 1)
Table 1
Pragmatic design features of the IMPAKT ERAS Trial
PRECIS domain(s)
|
Assessment
|
Eligibility criteria
|
This study will enroll all eligible major abdominal surgery patients within existing enhanced recovery pathways (colorectal, surgical oncology, & ventral hernia procedures) at a single academic medical center with few exclusions.
|
Recruitment path
|
The IRB has approved a modified research consent process for this study, which leverages a shortened consent form. Eligible patients will be recruited and consented by a physician member of the Perioperative Medicine team during the analgesic nerve block consent process.
|
Setting
|
Care will occur at a single, tertiary, academic medical center.
|
Organization intervention
|
Usual enhanced recovery pathway care will be administered for patients by clinical staff. A blinded study medication containing either ketamine or saline placebo will be initiated as a bolus at the induction of anesthesia and maintained as a postoperative infusion for up to 48 hours.
|
Experimental and comparison interventions – delivery
|
The OR pharmacy will prepare the blinded study medications and will oversee the randomization schedule in variable week blocks.
|
Experimental and comparison interventions – adherence
|
The lead study pharmacist reviews weekly reports of study drug administration to patients. Cross over events are identified by these reports. Adherence to the Enhanced Recovery Pathways is consistently monitored by a clinical dashboard that reports data in real-time. Early discontinuation of a study drug order requires providers to indicate a reason for cessation, which is captured in the EMR (e.g. side effects, early discharge anticipated).
|
Follow-up
|
In-hospital outcomes, including side effects and adverse outcomes, are monitored every 6 months by an independent Data Safety and Monitoring Board.
|
Primary trial outcome
|
All outcomes of this study are pragmatically designed to be readily obtained within the clinical EMR. The primary outcome is the hospital length of stay, defined as the time from start of surgery until hospital discharge. Secondary outcomes, including adverse perioperative events, opioid requirements, frequency of antiemetic administrations, and postoperative nasogastric tube placement are also pragmatically available through the EMR.
|
Analysis of primary outcome
|
All participants who are randomized will be included in the intention-to-treat statistical analysis, according to group assignment.
|
PRECIS, Pragmatic Explanatory Continuum Indicator Summary; IRB, Institutional Review Board; OR, operating room; EMR, electronic medical record
|
Interventions
In congruence with the clinical pharmacy’s current process of batching the production of ketamine infusions, randomization will occur in one-week cluster format, scheduled and tracked by the study pharmacist. All patients enrolled within a cluster will receive the same intervention (ketamine or saline placebo), with the choice being determined by a pre-generated random sequence. Both patients and their medical providers will be blinded to their randomization arm.
Participants assigned to the intervention arm will receive an intraoperative bolus of ketamine (0.5mg/kg) followed by an infusion (5 mcg/kg/min) until fascial closure. A postoperative ketamine infusion (2.5 mcg/kg/min) will begin in the post-anesthesia recovery unit and continue for up to 48 hours; however, it may be discontinued earlier for intolerable side effects, adverse events, or in preparation for patient discharge. Patients assigned to the placebo arm will receive an equivalent volume of saline for the bolus and infusions (Box 2). All other elements of the standard ERAS protocol will remain unchanged in both arms (Boxes 3 and 4). The Anesthesia Perioperative Consult Service rounds daily on all ERAS patients to provide clinical care under the guidance of established protocols, including an algorithmic and consistent approach to the treatment of breakthrough pain. This team is the primary service to address pain and nausea, and a member of this team is available in-house 24 hours per day, seven days per week (24/7).
Box 2: Ketamine and placebo interventions for the IMPAKT ERAS trial
Intravenous Ketamine Intervention
|
Intraoperative:
0.5 mg/kg bolus at anesthetic induction followed by intraoperative infusion of 5 mcg/kg/min until fascial closure, up to a maximum of 100 kg
Postoperative:
Infusion of 2.5 mcg/kg/min begun in PACU and continued for 48 hours postoperatively or until discontinued, up to a maximum of 100 kg
|
Placebo Intervention
|
Intraoperative:
Placebo normal saline bolus and infusion of an equivalent volume mirroring the ketamine intervention
Postoperative:
Placebo normal saline infusion of an equivalent volume mirroring the ketamine intervention, continued for 48 hours postoperatively or until discontinued
|
mg, milligrams; kg, kilogram; mcg, micrograms; min, minute; PACU, postanesthesia care unit
|
(Insert Box 3 here: longer than 1 page and placed at end of file per author instructions)
Box 4 Anti-emetic protocols for major intra-abdominal surgery ERAS care pathways included in the IMPAKT ERAS trial at Vanderbilt University Medical Center
ERAS Perioperative Anti-Emetic Regimen
|
Pre and Intraoperative Strategies:
• Number of PONV prophylaxis agents should equal the number of Apfel Risk Factors
• Avoid use of volatile agents: Total Intravenous Anesthesia (TIVA) with propofol is preferred
• Intraoperative Pharmacologic Agents:
Dexamethasone: 8mg IV after induction unless given in TAP blocks
Ondansetron: 4mg IV given 30 minutes prior to emergence
Haloperidol: 1mg IV given during skin closure, reduce to 0.5mg IV in patients > 65 years
Postoperative Pharmacologic Agents:
Ondansetron: 4mg IV/PO q 6 H PRN (write:1st line for nausea and vomiting)
Haloperidol: 0.5–1 mg IV PRN q 4–6 H PRN (write: 2nd line after ondansetron)
Scopolamine patch (write: 3rd line option only if active PONV despite above)
Promethazine: 6.25–12.5 mg IV/PO q 4 H PRN (write: 4th line option)
|
ERAS, enhanced recovery after surgery; PONV, postoperative nausea and vomiting; mg, milligrams; IV, intravenous; TAP, transversus abdominis plane; PO, per os; q, quaque (every); h, hours; PRN, pro re nata (as needed)
|
Outcomes
Hospital LOS, defined as the time between anesthesia start and patient discharge, was chosen as the primary outcome because it is an objective marker of functional recovery after surgery. It is the customary outcome measured when evaluating the overall efficacy of ERAS protocols to facilitate faster recovery compared to routine care. Moreover, LOS represents a metric that is important to patients and families, physicians, and healthcare systems. Secondary outcomes include the incidence of adverse perioperative events, such as rapid response activation and escalation of care to an ICU. We will also investigate the frequency of antiemetic administrations and placements of postoperative nasogastric tubes for severe ileus.
The analgesic impact on total opioid consumption, measured as morphine milligram equivalents of opioid use until hospital discharge, will be investigated as a secondary outcome, as opioid minimization has been identified as an important contributor to the success of an ERAS program. Within our institution’s abdominal ERAS programs, perioperative median morphine milligram equivalents (MMEs) are intentionally minimized. Therefore, it is possible that a further reduction in opioids may of be negligible benefit, and perhaps the side effect profile of ketamine impairs recovery rather than promotes it. The incidence of side effects will be compared across the two arms by examining the documented reasons for and frequencies with which the investigational drug infusions are discontinued early, prior to the 48-hour duration as intended.
Sample size planning
ERAS pathways for major abdominal surgery were established at VUMC in 2013. Prior to the addition of perioperative ketamine infusions to these protocols in 2016, the median hospital length of stay was 4.3(5.3) days. To determine the sample size needed for sufficient power to detect a meaningful difference, the sample size calculation assumed 85% power, a two-sided type I error rate of 5%, and a 10%, reduction in hospital LOS. The calculation estimated the within period correlation from preliminary data concerning the LOS for major abdominal surgeries to be about 0.014. Given these assumptions, about 757 patients would need to be included in each arm. An increase in sample size was deemed necessary to accommodate for the two post-randomization exclusions that are pre-specified. Accordingly, the total target enrollment will be 1,544 participants over a 2 year period.
Recruitment
Adult patients (≥ 18 years) presenting to VUMC on a weekday for an elective, major abdominal surgery within an established ERAS protocol are eligible for enrollment. Per usual care, patients are approached on the day of their surgery by a physician member of the Anesthesia Perioperative Consult Service who describes our care goals and multimodal pathways, and then obtains consent to perform a regional or neuraxial nerve block. Regional blocks are offered for laparoscopic and small open incisions (< 10 cm) and include abdominal wall blocks (e.g. rectus sheath, transversus abdominis plane). Neuraxial anesthesia (thoracic epidural catheter) is offered for patients planned to have open procedures. The IRB approved a modified research consent process for the trial with research consent also obtained when clinical consent for the regional or neuraxial nerve block is obtained. If the patient elects to participate, written consent will be collected by the Anesthesia Perioperative Consult Service physician and uploaded to the EMR.
Allocation
Randomization will occur in one-week cluster format, with the treatment arm assigned by a pre-generated random sequence allocated by the study pharmacist. Weekly batches of the appropriate study drug are prepared by the Operating Room pharmacy in generic infusion bags with a non-identifying study label. The study drug is stocked in a refrigerated Omnicell located in preoperative holding, and the contents are replaced with the new batch each week. The lead study pharmacist reviews weekly reports of study drug dispersal and administration to identify cross over events.
Data analysis and management
Initial analysis will use descriptive statistics and data visualizations to both identify and address spurious values, and to characterize the study cohort. Characteristics will be described overall and grouped by study arm, without a plan to compare the study arms with statistical tests. Categorical variables will be described using frequencies and proportions; continuous variables will be described using means and standard deviations as well as medians and interquartile ranges (IQR). Missingness will be recorded for each variable.
The primary outcome is hospital LOS, measured in days. It is expected that LOS will have a skewed distribution, and thus non-parametric methods are preferred. However, it is not possible to consider the clustering or baseline covariates using a Wilcoxon test. Therefore, a proportional odds regression model will be used. The model will include baseline characteristics such as age, body mass index (BMI), smoking status, opioid use, and type of surgical procedure. Multiple imputation based on predictive mean matching will be used for missing covariates. Continuous variables will be modeled flexibly using cubic splines. Clustering will be taken into account using a mixed-effects model. If there are patients included twice because they underwent two separate hospitalizations for elective major abdominal surgeries, those patients will also be included as a random effect. The common odds ratio from the model will be estimated; this quantity can be interpreted similar to the concordance probability and is robust to the proportional odds assumption. Model-assisted estimates of mean and median length of stay will also be reported. Bootstrapping will be used to generate appropriate confidence intervals.
The primary analysis will be an intention to treat analysis including all randomized participants analyzed according to group assignment. If a participant or provider withdraws the participant from the study following study drug administration, the participant will receive care at the provider’s discretion, but will remain in the intention to treat statistical analysis.
The secondary outcomes will be explored in a similar way. The proportional odds model will be used to evaluate differences in opioid use in the hospital. Occurrence of events will be reported as rates with confidence intervals. Logistic regression will be used to compare the odds of events between groups.
In all statistical modeling, emphasis will be placed on effect sizes over p-values. In addition, continuous variables will be modelled flexibly using cubic splines. Differential treatment effects will be explored by examining the interaction between the treatment indicator and the putative subgrouping variable. Effects of treatment by sex will be analyzed. As before, continuous variables will not be categorized, and any interactions will be evaluated as continuous variables. Subgroup analyses will not occur in reporting the main results of this trial unless there is evidence of a differential treatment effect on interaction testing.
Monitoring
Patients enrolled in the trial are closely followed by their surgical services and the Anesthesia Perioperative Consult Service, both of which round daily and have providers available in the hospital at all times. Any deviations from the protocol or unanticipated events are identified promptly and immediately reported to the Principal Investigator (PI), who is responsible for reporting the events to the IRB.
The ERAS protocols at VUMC provide guidelines for the clinical management of acute perioperative pain. Adherence to the protocols is vital to the integrity of the trial, as all elements of the pathway (apart from ketamine administration) are to remain comparable between groups. A dedicated Anesthesia Perioperative Consult Service offers consistency in applying these ERAS principles. Furthermore, adherence to ERAS elements is regularly monitored through an electronic dashboard that provides real-time analysis of patient data and ordering practices (Figs. 1, 2).
Data and safety monitoring
A Data Safety and Monitoring Board (DSMB) of three independent, unaffiliated, multidisciplinary specialists meet at least every 6 months to ensure that the rights and well-being of participants are protected and the reported data are accurate and complete. At each meeting, DSMB members are apprised of the accrual of subjects, fidelity to the protocol, and reports of serious adverse events and unanticipated adverse events. The DSMB is also unblinded to the results of the outcomes at their meetings during a closed session. The DSMB provides advisory council to the PI and has authority to recommend halting the trial if continuation is deemed unsafe. The final recommendations of the DSMB will be immediately reported to the IRB.
Ethics
Ketamine is an analgesic that has been proven to reduce opioid consumption after surgery. It also has a role in reducing the development of chronic pain. Compared to opioids, ketamine has a favorable side-effect profile, including the maintenance of respiratory and cardiovascular stability. Ketamine does not act at the mu opioid receptor, and therefore is not associated with nausea, itching, constipation, and tolerance. We have been safely utilizing perioperative ketamine at VUMC for several years as part of our enhanced recovery pathways. The IMPAKT ERAS study design does not involve the introduction of a new treatment or intervention, but rather a structured, yet pragmatic, investigation of the impact of currently administered medical care, with an intent to leverage data that is already collected as part of routine clinical care to use as outcomes. The IMPAKT ERAS trial has been deemed minimal risk above clinical risk by the IRB, allowing for alteration of the informed consent process.
Adverse reactions to ketamine
Adverse reactions and allergies to ketamine are routinely reviewed by both medical providers and pharmacy personnel prior to ordering and administering the drug. Ketamine is a controlled substance and is highly regulated, recorded, and tracked. Patients are evaluated daily at the bedside by our Anesthesia Perioperative Consult Service, and patients are screened for side effects of ketamine on daily rounds. Nurses are trained to immediately report any suspicion of an adverse ketamine reaction to a member of our team, who is available in the hospital at all times.
Selection of patients
Only patients receiving care through our enhanced recovery pathway (which includes the standard administration of ketamine) whose provider agrees is appropriate for inclusion and who signs the research consent will be eligible for participation.
Ineffective pain control
Patients with breakthrough pain will receive routine treatment, including opiate administration if needed, at the discretion of medical providers through the guidance of our analgesic protocols.
Breach of confidentiality
One of the risks specific to the research study is potential breach of confidentiality. The investigators have access to the medical data as part of their job standing. To minimize this risk, all database work will be performed on VUMC approved and password protected servers that are physically located in the VUMC computing environment and maintained by VUMC security standards. Only the PI and key study personnel will have access to study specific non de-identified data.
Dissemination
The results of the IMPAKT ERAS trial will be submitted for presentation at local, national, and international medical conferences, as well as academic institutions through visiting and invited lectureships. Furthermore, we aim to publish our results in a high-impact, scientific journal that will distribute the data to a broad clinical audience.