The Ketamine for Acute Pain after Trauma (KAPT) trial is a single-center, parallel-group, randomized, controlled comparative effectiveness trial of our MMPR to our MMPR plus a sub-dissociative ketamine infusion for the first 72 hours after admission. We hypothesize that the addition of ketamine will result in lower opioid exposure in injured patients compared to our generic MMPR alone as evidenced by lower average oral MME per day.
Trial Design:
The trial is registered at ClinicalTrials.gov (Identifier: NCT04129086) and was designed in accordance with the Standard Protocol Items: Recommendations for Interventional Trials 2013 guidelines. (Figure 1) This manuscript with adherence to SPIRIT reporting guidelines.11
Study Setting:
This study is being conducted at the RDTI at the Memorial Hermann -Texas Medical Center, which is one of two level 1 trauma centers in Houston, Texas. It is a high-volume, teaching hospital for the McGovern Medical School at the University of Texas Health Science Center at Houston.
Eligibility Criteria
All adult patients (≥16 years of age) admitted to the adult trauma service at the intermediate or intensive level of care are screened for eligibility. Pregnant women, prisoners, and those not expected to survive are excluded. Additionally, patients with the following contraindications to ketamine are excluded: allergy to ketamine, poorly controlled hypertension, cardiac arrhythmia (including atrial fibrillation), congestive heart failure, unstable coronary artery disease or recent myocardial infarction, cirrhosis, dementia, movement disorder (e.g. Parkinson’s), or seizure disorder. Additionally, for those unable to provide past medical history, patients are excluded if they have arrhythmia on EKG, are older than 65 years, present after fall from standing, or have median sternotomy scar.
Informed Consent
We obtained a waiver of consent to randomize patients due to the following:
- The intervention is minimal risk; the ketamine dosing proposed is already being used by prehospital personnel, emergency room physicians, and trauma surgeons at similar or higher doses.
- The research could not be practicably carried out without the waiver; due to the acute clinical status of the trauma patient population (intubation, intoxication, severe pain, emotional stress), truly informed consent cannot be obtained in the vast majority of patients (or their legally authorized representative ) before the ketamine would be given.
- The waiver would not adversely affect the rights and welfare of the subjects; patients often receive opioids and/or ketamine in the Emergency Department acutely without their knowledge or written consent, they simply want pain control. We will be providing pain control and rescue opioids would not be withheld from the intervention group.
Once the patient is randomized, a member of the trauma research team attempts to contact either the patient or legally authorized representative to obtain consent for using Protected Health Information, to return to perform discharge surveys, and to contact after discharge for post-discharge surveys. Consents are obtained by trained research personnel and patients are given written study information. If, after 5 days, the patient remains unable to self-consent and no LAR is available to consent, the consent is waived and data included. Additionally, if the subject does not survive following the traumatic injury or is discharged from the hospital before the study team is able to obtain consent, their information is included in data analysis.
Interventions:
Usual Care Group:
Trauma patients at the RDTI are initially prescribed with an opioid-minimizing MMPR informed by the MAST trial.3 This regimen consists of four different classes of medications given in a scheduled fashion: acetaminophen, naproxen, gabapentin, and lidocaine patch(s). Tramadol or oxycodone are administered as needed for additional pain control. The bedside provider is allowed to adjust the pain regimen as needed.
Ketamine Plus Usual Care Group:
Patients in the ketamine group receive sub-dissociative ketamine for a maximum of 72 hours in addition to MMPR. (Figure 2) Ketamine is administered as a bolus dose of 0.35mg/kg and then given as a continuous infusion of 0.1-0.25 mg/kg/hr. This dosing was determined based on the American Society of Regional Anesthesia and Pain Medicine, the American Academy of Pain Medicine, and the American Society of Anesthesiologists consensus guidelines.12 Ideally, the infusion is continued for a minimum of 24 hours and a maximum of 72 hours, however the bedside clinician is free to alter the pain regimen as seen fit.
The treatment is initiated in the emergency department. In both treatment groups of this pragmatic trial, adjustments may include withholding or adjusting the dosage of MMPR medications due to comorbidities such as kidney or liver disease; withholding a medication due to interactions with other medications; and de-escalation of medications once adequate pain control is achieved. In both regimens, oral medications are also available as needed for breakthrough (moderate and severe) pain—oxycodone and/or tramadol. Although their use is discouraged in general clinical practice at our institution, intravenous opioids can be administered as needed.
Outcomes
The primary outcome is opioid exposure defined as average MME per patient per day. MME/day was calculated by converting all sources of opioids during the hospital stay, including emergency department and operating room, to a single MME value using a standardized conversion chart and dividing by length of hospital stay. (Table 1MME/day was chosen as the value can be easily calculated by other centers with which to compare their patients’ opioid exposure and the value accounts for length of stay.
Secondary opioid exposure outcomes include total MME and opioid prescribing at discharge. Other secondary outcomes include pain scores (Behavioral Pain Score (BPS), Numeric Rating Score (NRS), and/or Defense and Veterans Pain Rating Scale (DVPRS), potential opioid-related complications (delirium, unplanned intubation, and unplanned admission to an intensive care unit), and ketamine-related information (time to infusion initiation, duration of infusion, etiology for infusion cessation). Additional secondary outcomes include use of regional anesthesia, lengths of stay, and hospital costs. Patient requests to discontinue ketamine infusions and incidence of delirium are also being recorded.
At 6-month follow up, patients are contacted to assess for the presence of persistent post-traumatic pain and persistent opioid use. Additionally, at that time, patient-centered outcomes including quality of life will be assessed using Euro-QOL EQ-5D-3L .13
Sample Size
Due to the uncertainty of an exact treatment effect of ketamine in our patient population and the potential for contamination of the control group if the trial is continued for a prolonged length of time, we plan to perform the largest feasible study over 18 months. Over a sample period of 6 months (November 2018 – April 2019), manual review of admission to the trauma service revealed 302 potentially eligible patients. Extrapolating this data, we expect 906 potentially eligible patients over the 18-month trial period. Given the emergent nature of the intervention, general difficulties enrolling severely injured patients, and several concomitant randomized clinical trials, we conservatively estimate enrolling 40% of potentially eligible patients.
We anticipate recruiting approximately 395 patients into the current trial. Based on its skewed distribution, MME’s per day are postulated to be Gamma distributed (MME’s need not be integer values). Using the 50th and 75th percentiles, MME = 45 and 67 respectively, from our previous data we estimate the control condition as following the distribution ~Gamma(shape = 2.7784923, scale = 18.343307). Further, assuming that the ketamine condition results in a 20% decrease in MME’s (50th and 75th percentiles MME = 36 and 53.6 respectively) we assume the observed data for this condition is distributed ~Gamma(shape = 2.7784923, scale = 14.674646). We assume that clustering within unit induces a substantial intracluster correlation (ICC) = 0.2. Finally, we stipulated that a 75% chance that the ketamine treatment reduces MME’s by at least 15% relative to treatment as usual would constitute sufficient evidence to warrant a larger clinical trial. Assuming the previously stated sample size, effect, ICC and decision rule, K = 1000 Monte Carlo simulations suggest there is a > 99% chance of observing this effect.
Sequence Generation
Randomization is performed at admission while in the emergency department in a 1:1 allocation and stratified by unit of admission. The random sequence was generated by an independent statistician. Allocation is performed by the in-house research assistant using REDCap database only accessible by investigators. This database is password protected and will not be shared with those outside of study. 14
Concealment Mechanism
Providers and patients are not to be blinded to treatment allocation due to feasibility and cost. To address this issue, the opioid administration and pain score data (entered by nurses during routine clinical care) will be captured directly from the electronic medical record and the majority of outcomes will be obtained using our trauma registry (blinded outcome assessor).
Analysis
The data analytic strategy will use Bayesian inference, applying generalized linear multilevel modeling with level-two random effects or fixed effects (depending on model convergence) to account for clustering of participants within department and, where applicable, observations within participants. Modeling will use R v. 3.4 and Stan v. 1.10.15,16. For the purposes of evaluating the comparability of groups, a posterior probability that a difference/correlation exists of > 95% will constitute evidence for statistically reliable differences. If potential confounders are identified we will perform two sets of analyses: one in which the confounder(s) is included as a covariate and one in which it is not.17,18 This will permit determination of the degree to which any group differences might confound conclusions regarding treatment. All analyses will use intention-to-treat principles. Bayesian approaches will implement joint modeling of observed outcomes and the missing data which is robust to ignorable missingness (missing completely at random and missing at random)19 Sensitivity analyses will evaluate robustness of analytic conclusions to missing data. Non-ignorable missing data patterns will be addressed through pattern-mixture modeling methods.20 Convergence of Bayesian analyses on the posterior distributions via Monte-Carlo Markov chain (MCMC) will be assessed via graphical (Gelman-Rubin Plots) and quantitative (Gelman-Rubin Diagnostics and Effective Sample Size) evidence. Evaluation of posterior distributions will permit statements regarding the probability that effects of varying magnitudes exist, given the data. Diffuse, neutral priors will reflect the initial uncertainty regarding effect sizes. For all generalized linear multilevel models, priors for regression coefficients will be specified as ~Normal (µ=0, σ2=1000) on the identity or log-scale depending upon the model, level one error variances will be specified as ~Half-T (df = 3, mean = 0, standard deviation = 100). Prior distribution for level two variances will use ~Half-T (df = 3, mean = 0, standard deviation = 1000). Priors for the comparison of proportions will be specified as ~Beta (α=0.5, β=0.5). For all subgroup analyses using multilevel models the approach will follow that used in Tyson, et al. 21
Data and Safety Monitoring Board
As this is a comparative effectiveness trial utilizing ketamine at a lower dose than commonly used in the Emergency Department for moderate sedation during procedures (fracture reduction and splinting, tube thoracostomy, etc.), no data and safety monitoring board will be formed. Major adverse events, including death, will be reported to the Institutional Review Board (IRB).
Trial Status
The IRB of McGovern Medical School approved the study protocol on September 30, 2019. Enrollment was delayed by the COVID-19 pandemic but began on July 9th, 2020 and is scheduled to continue for 18 months. As of April 23, 2021,193 patients have been enrolled.