Introduction
Pain is the most common complaint of trauma patients attending the emergency department (ED), with prevalence of 70% reported in the prehospital setting and 91% in EDs [
1,
2]. However, undertreatment of pain (oligoanalgesia) remains a widespread problem in the emergency setting [
3‐
6]. Various reports have attributed oligoanalgesia to failure to assess/underestimation of pain by healthcare workers [
7‐
9], lack of national/institutional guidelines for pain management [
8‐
10] and limitations of currently available therapies [
9].
Treatment options for acute pain management include paracetamol, non-steroidal anti-inflammatory drugs (NSAIDs) and weak opioids for mild to moderate pain, and strong opioid analgesics for severe pain [
11], although there is considerable heterogeneity in the therapeutic approach across Europe. Paracetamol and NSAIDs are weak analgesics with slow onset of action when administered orally, and risk of overdose if the patient has previously self-medicated. Furthermore, NSAIDs can be associated with gastrointestinal problems, nephropathy, cardiovascular disorders, and reduced fracture healing. Strong opioid analgesics are controlled drugs that require patient monitoring for potential side effects such as respiratory and central nervous system depression, nausea and vomiting. Intravenously administered (IV) analgesia, while aiding onset of action, requires additional healthcare resource, can be distressing for the patient and may be difficult in emergency rescue situations. There is rising use of alternative non-opioid and multimodal analgesic therapies such as subdissociative-dose ketamine and nitrous oxide [
12]. Nitrous oxide (50:50 with oxygen) has the benefit of being non-invasive and non-narcotic with quick onset of action, rapid reversibility and few side effects [
13‐
15], but the gas canisters are cumbersome and impractical when attending emergency situations in the field [
16] and it should not be used if the patient has any condition where air is trapped in the body and expansion would be dangerous [
17]. Thus, there is an unmet need for a rapid-acting, safe, effective and easy-to-use treatment for trauma pain in the emergency setting.
Methoxyflurane is a volatile inhalational analgesic that provides rapid short-term analgesia using a portable hand-held inhaler device (Penthrox
®, 3 mL dose; Medical Developments International, Scoresby, Australia) and may provide an effective non-narcotic option for emergency pain management in the prehospital and ED setting. Methoxyflurane is self-administered by the patient, has a rapid onset of action (within around 4 min or 6–10 inhalations [
18,
19]), its effects are quickly reversed (within 3–20 min after inhalation stops) and there are no reported drug interactions at analgesic doses [
19]. These characteristics make methoxyflurane suitable for use as a sole agent, or as a bridging agent to other analgesia, and its use may have an opioid-sparing effect. One inhaler (3 mL methoxyflurane) provides 25–30 min of analgesia with continuous inhalation [
19]; or approximately 1 h of analgesia under intermittent inhalation conditions [
20]. Low-dose methoxyflurane analgesia has a well-established safety profile: no respiratory depression or clinically significant effects on vital signs have been reported and adverse events (AEs) are usually transient and self-limiting [
18‐
21].
Methoxyflurane has been used extensively in Australia and New Zealand as emergency and procedural analgesia for more than 40 years [
21‐
23], and is now licensed in Europe and other territories for the emergency relief of moderate-to-severe pain associated with trauma in conscious adult patients [
24]. Its use as emergency analgesia is supported by robust data in the literature [
21‐
23]. To date, key data in Europe are from STOP!, a randomised, double-blind, placebo-controlled UK study in 300 patients with acute trauma pain (baseline score of 4–7 on the 11-point numeric rating scale, NRS) [
18,
25,
26]. STOP! showed a significantly greater reduction in visual analogue scale (VAS) pain scores overall for methoxyflurane vs. placebo (− 30.2 vs. − 15.2 mm,
p < 0.0001) with a median time to first pain relief of 4 vs. 10 min [
18]. More recently, the InMEDIATE study demonstrated a significantly greater decrease in NRS pain intensity over the first 20 min of treatment for methoxyflurane vs. standard analgesic treatment chosen by the treating physician according to local protocols (− 2.5 vs. − 1.4,
p < 0.001), with a median time to first pain relief of 3 vs. 10 min (
p < 0.001) in 305 trauma patients in Spanish EDs [
18]. Both studies showed high patient and healthcare professional (HCP) satisfaction with methoxyflurane treatment [
18]. However, additional comparative studies versus the best currently available therapeutic options are needed to fully elucidate the place of methoxyflurane in emergency analgesia, including in extreme conditions such as helicopter rescue, which is the subject of an ongoing interventional study [
27].
The aim of the present study was to compare the short-term pain relief (up to 30 min after randomisation) and safety with inhaled methoxyflurane versus standard analgesic treatment (SAT) in adult patients presenting to Italian emergency medical centres with acute trauma-related pain. SAT comprised IV morphine for severe pain (NRS ≥ 7) and IV paracetamol or IV ketoprofen for moderate pain (NRS 4–6), consistent with Italian recommendations on pain management in the emergency setting [
28].
Methods
Study Design
Methoxyflurane in Emergency Department in ITAly (MEDITA) was a phase IIIb, randomised, active-controlled, parallel-group, open-label study conducted at 15 emergency medical centres (both prehospital and ED settings) in Italy. Adult patients with moderate-to-severe trauma-related pain were randomised 1:1 to receive methoxyflurane or SAT. All study procedures were performed on the day of enrolment, with a follow-up telephone call 14 ± 2 days later to record any AEs and concomitant therapies. The full methodology for this study has previously been reported [
29‐
31].
The study was approved by the Italian Medicines Agency (AIFA). The co-ordinating ethics committee, Comitato Etico Regione Toscana—Area Vasta Centro, Florence, Italy, approved the trial protocol on 1 December 2017. In addition, all study documents and procedures were reviewed and approved by the appropriate ethics committees at each centre (see Table S2 in the electronic supplementary material for list of all ethics committees). All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards. Informed consent was obtained from all individual participants included in the study. Given the emergency setting and the requirement for rapid analgesia, if the patient was unable to provide written informed consent, witnessed verbal consent was obtained, with the patient signing the informed consent as soon as they were able.
Participants
Patients aged at least 18 years with trauma (fracture, dislocation, crushing, contusion) to a single limb and an NRS pain score ≥ 4 presenting at the hospital for triage or rescued in the prehospital environment through the Italian emergency medical service were eligible for the study. Patients had to be medically stable, alert and collaborative, and able to communicate with the investigator to perform the study activities, including providing informed consent. Exclusion criteria included ongoing analgesic treatment for chronic pain or use of any other analgesic in the previous 5 h (8 h for diclofenac), pregnancy or lactation, dynamics of at-risk trauma, and contraindications to methoxyflurane administration as per the Summary of Product Characteristics (SPC) [
19] or to any of the SAT. Full inclusion and exclusion criteria have previously been reported [
29].
Randomisation and Blinding
Patients were randomised to study treatment (methoxyflurane or SAT) via a centralised Interactive Web Response System that guaranteed allocation concealment, set up within the electronic case report form (eCRF). Randomisation was in a block randomisation scheme of four without stratification. As a result of the different routes of administration, need for rapid analgesia and characteristic fruity smell of methoxyflurane, study treatment was administered in an open-label fashion, commencing as soon as possible after randomisation.
Description of the Interventions
Patients randomised to methoxyflurane received one hand-held inhaler with a 3-mL vial of methoxyflurane. Following patient allocation to the experimental group, the methoxyflurane inhaler (Penthrox) was prepared by trained study HCP by adding the methoxyflurane liquid in the vial to the inhaler via a one-way valve, where it is absorbed by a polypropylene wick. Once absorbed, the liquid vaporizes and the patient inhales the vapour through the mouthpiece. Methoxyflurane was self-administered by the patient under the supervision of a trained study HCP. The patient was instructed to inhale intermittently from the device and could control their own level of analgesia by inhaling more or less frequently, or covering the diluter hole at the mouthpiece end with their index finger to obtain greater analgesia. The patient was also instructed to exhale back into the mouthpiece, so that any exhaled methoxyflurane was adsorbed by the activated charcoal chamber, preventing release of methoxyflurane in the vicinity of the patient. One inhaler (3 mL methoxyflurane) was expected to provide approximately 1 h of analgesia under suggested intermittent inhalation conditions, or 25 min with continuous inhalation [
19,
20].
Patients randomised to SAT received analgesic medications that currently comprise standard analgesic treatment in Italy, determined on the basis of the intensity of the patient’s pain. Patients with severe pain (NRS ≥ 7) received IV morphine (0.10 mg/kg). Patients with moderate pain (NRS 4–6) received IV paracetamol (1 g) or IV ketoprofen (100 mg), chosen by the investigator on the basis of availability and local practice, and considering any prior history of allergy in the patient. The IV treatment was diluted and infused over 10 min as soon as possible after allocation of the patient to the control group. Venous access was obtained prior to randomization according to local clinical practice.
Rescue medication was permitted for patients in either treatment group from 25 min after randomisation (after measurement of pain relief for this time point), or earlier at the investigator’s discretion if the patient’s pain worsened, or improvement was insufficient. Rescue medication was administered according to local practice.
Outcomes
The primary outcome was the change in VAS pain intensity from baseline (time of randomisation) to 3, 5 and 10 min, with a secondary outcome of the change in VAS pain intensity from baseline to 15, 20, 25 and 30 min. Pain intensity was measured by asking the patient to respond to the question “How much pain do you feel at this moment?” by marking a 0–100 mm VAS, where 0 = no pain and 100 = maximum pain. Pain intensity assessments ceased if rescue medication was administered. Given the emergency/rescue setting and the acute nature of the endpoint, a specially trained HCP could assist the patient in completing the VAS, if required, in which case the patient authenticated the recording with a signature and date as soon as they were able. While the NRS was considered adequate for enrolment, VAS pain intensity was chosen as the primary endpoint because it is evaluated on a continuous scale, thus increasing the sensitivity and power of the study.
Further secondary outcomes were the use of rescue medication, the time of onset of pain relief as subjectively reported by the patient, ratings of study treatment efficacy by the patient, and ratings of the practicality of study treatment by the treating HCP. The patient rated the overall efficacy of the study treatment and the HCP rated the practicality of using the study treatment on a 5-point Likert qualitative scale (“Poor”, “Fair”, “Good”, “Very Good” or “Excellent”) at 30 min after randomisation. An exploratory outcome was the proportion of patients in the methoxyflurane group resorting to closure of the diluter hole (to administer a higher concentration of methoxyflurane).
Secondary safety outcomes were the incidence of AEs, and vital sign measurements at baseline and 10 and 30 min after randomisation. AEs (not related to the trauma presentation) were recorded from the time of randomisation until the safety follow-up telephone call at 14 ± 2 days after treatment. The final diagnosis of the trauma category (fracture, dislocation, crushing, contusion) was also recorded during the safety follow-up telephone call. All data collected during the trial were entered into an eCRF system accessible via the internet and fully monitored by clinical research associates to ensure data quality; in addition, the three highest-recruiting centres underwent external audit.
Objectives
The primary objective of the trial was to demonstrate non-inferiority of methoxyflurane compared to SAT in Italy (IV morphine/ketoprofen/paracetamol) for the treatment of moderate-to-severe acute pain in terms of the primary outcome in all study patients, with a secondary objective to demonstrate superiority of methoxyflurane compared to SAT for the same endpoint. The co-primary objective was to demonstrate superiority of methoxyflurane compared to SAT (IV ketoprofen/paracetamol) for the treatment of moderate acute pain in terms of the primary outcome in patients with baseline NRS 4–6. All other study objectives are detailed in the protocol publication [
29].
Sample Size
A sample size of 108 patients per treatment group was estimated to provide 90% power to determine non-inferiority of methoxyflurane versus SAT for the primary outcome measure, assuming a non-inferiority margin of 1.0, a standard deviation of 2.5 [
18] and a significance level of 0.05. Allowing for 20% of patients being non-evaluable, it was planned to randomise a total of 136 patients per treatment group. The co-primary objective was not considered in the sample size calculation because enrolment was not balanced by baseline pain severity; therefore, the number of patients enrolling with moderate pain was unpredictable.
Statistical Analysis
The primary outcome measure was analysed using a linear mixed-effect model for repeated measures adjusted for baseline VAS score, and the interaction between time point and treatment. The statistical model was used to calculate the treatment difference (methoxyflurane − SAT) and associated 95% confidence interval (CI) at 3, 5 and 10 min for all patients, and patients with moderate pain (NRS 4–6) at baseline. The primary analysis was the overall test for treatment effect considering all three time points; non-inferiority was to be concluded if the estimated upper 95% confidence limit was below 1.0 and superiority was to be concluded if the estimated upper 95% confidence limit was below 0. The mean change from baseline with 95% CI was estimated for each group at 15, 20, 25 and 30 min after randomisation (secondary outcome).
It was planned to compare the percentage of patients who resorted to rescue medication between the treatment groups using a Z test; however, as a result of the low number of patients resorting to additional analgesia, the non-parametric Fisher’s exact test was used. Kaplan–Meier curves were used to present the time from randomisation to the onset of pain relief for each treatment group and, if appropriate, the treatment groups were compared using a Cox proportional hazards model. For patient efficacy ratings and HCP practicality ratings, the percentage of patients in each response category was compared between the treatment groups using a Mann–Whitney U test. The proportion of patients in the methoxyflurane treatment group who resorted to closure of the diluter hole was evaluated as a relative frequency with 95% CI (Clopper–Pearson).
AEs were coded using the Medical Dictionary for Regulatory Activities version 22.0 and summarised descriptively (using absolute frequencies and percentages) for each treatment group. All efficacy and safety analyses were performed using the intention-to-treat (ITT) population. No imputation of missing data was performed; therefore, only patients with available data were considered for each analysis. Randomisation, or the time at which treatment was assigned, represents the baseline time to which all the times of observation for the study endpoints refer.
Discussion
This trial demonstrated both non-inferiority and superiority of methoxyflurane compared to SAT in terms of the decrease in pain intensity achieved in the first 10 min post baseline. Methoxyflurane was superior to standard treatment for both moderate pain (IV paracetamol or ketoprofen) and severe pain (IV morphine). Given the intravenous administration of the comparator treatment, the change in the first 10 min post baseline reflected the most challenging comparison for methoxyflurane, and was directly relevant to the indication for methoxyflurane of emergency relief of moderate-to-severe trauma pain [
19]. Previous studies have shown that methoxyflurane is effective at the 20-min time point [
18,
25,
26,
32], thus the 10-min endpoint utilized in this study provides additional information about the efficacy profile of inhaled methoxyflurane. The secondary analysis of the change in VAS pain intensity to later time points also showed significantly larger reductions in pain intensity for methoxyflurane versus SAT up to 25 min. Although in the present study patients were provided with only one vial (3 mL) of methoxyflurane, ensuring approximately 25–60 min of effect, depending on frequency of inhalation [
19,
20], in clinical practice a second 3-mL dose could be administered if a longer analgesic effect is required [
19]. The absolute mean change from baseline in VAS pain intensity at 10 min for methoxyflurane in this study (− 23 mm from a baseline mean of 67 mm) is likely to represent a clinically meaningful change according to previous estimates of minimum clinically important difference in acute pain of approximately 20% reduction on the NRS [
33] or approximately 13 mm on the VAS [
34,
35].
The median time from randomisation to onset of pain relief was significantly shorter for methoxyflurane (9 min) compared with SAT (15 min) in this trial. This result was confirmed by the Gehan–Breslow–Wilcoxon test comparing the Kaplan–Meier curves. Given that the standard analgesic treatments were all administered intravenously, this may reflect the time taken to administer SAT as well as the onset of analgesic action. The time of randomisation was chosen as baseline with the specific intention of investigating not only the intrinsic efficacy but also the speed of drug administration in an emergency context, characterized by the need to intervene as quickly as possible to relieve the patient’s pain while being able to continue with the diagnostic-therapeutic process. The InMEDIATE study reported that the median time to onset of pain relief from the start of treatment was significantly shorter for methoxyflurane (3 min) than SAT (10 min), suggesting that the quicker onset of action of methoxyflurane in the present study is not solely due to the difference in time taken to administer the treatment [
32]. Similar to the present study, most patients in the SAT group in the InMEDIATE study received IV analgesics, although the majority of patients with severe pain received first-step analgesics rather than opioids. Median time to onset of pain relief for methoxyflurane in the placebo-controlled UK STOP! study was 4 min [
18] (5 min in the adult subgroup [
25]).
The effectiveness as a short-term treatment option for acute trauma pain, the ease of preparation and administration of methoxyflurane, and the fact that patients can self-administer treatment and control their own level of analgesia have the advantage of lightening the care burden in the ED compared with IV treatments. This is supported by the HCP ratings of the practicality of using study treatment, which was rated as “Excellent”, “Very Good” or “Good” in 90.3% of cases for methoxyflurane and 64.4% for SAT. Methoxyflurane does not require physiological monitoring during use, unlike opioid analgesics, and does not have the associated administration costs of IV morphine, which have been estimated at €14–22 in terms of workforce costs alone, rising considerably when costs of morphine- and IV-related AEs are included [
36]. Inhaled methoxyflurane may also be less stressful for the patient than intravenous administration, and more practical in the out-of-hospital environment.
Although more AEs were reported in the methoxyflurane group (17%) compared with the SAT group (3%), most of them were minor and transient, and only eight patients (5.9%) in the methoxyflurane group discontinued treatment because of AEs. The STOP! and InMEDIATE studies also reported higher AE rates for methoxyflurane vs. placebo and active treatment, respectively, but patient-reported outcomes in these studies indicated high patient acceptance of treatment. Indeed, 78% of adult patients in STOP! rated methoxyflurane treatment as excellent, very good or good [
25], and patients in InMEDIATE rated methoxyflurane a median of 9 out of 10 for pain control, comfort of treatment and safety (AEs) [
32]. No effects of methoxyflurane on vital sign parameters were observed, consistent with reports from other clinical trials in Europe [
18,
25,
26,
32] and findings of two observational studies including 1217 patients treated with methoxyflurane in the prehospital setting in Australia [
20,
37]. No cases of methoxyflurane abuse were identified in our study.
Strengths and Limitations
The study strengths include the multicentre, randomised design, allocation concealment (reflected by the homogeneous baseline data) and comparison with an active comparator treatment that reflects current practice in Italy [
28]. However, national- and local-level variations in pain management guidelines may limit the generalizability of the results to some other countries, and data from ongoing and recently completed studies in other countries are awaited [
38,
39]. The current study results are robust; consistent results were obtained for the primary outcome measure across analyses of all patients, patients with moderate pain treated with IV paracetamol/ketoprofen, and patients with severe pain treated with IV morphine. Secondary efficacy results for the onset of pain relief and patient ratings of efficacy also supported the conclusion of improved efficacy with methoxyflurane versus SAT. The study population included a broad range of adult patients with acute trauma pain in terms of age, gender, pain severity and injury type, reflecting a “real-world” emergency setting.
A limitation of this trial is the open nature of treatment administration. Although this design presents an intrinsic bias, it was considered the only practical option due to the nature of the trial setting and of the treatments being studied. A double-blind study, while ideal from a methodological perspective, would not have been feasible because of the different routes of administering the two treatments, unless using a double-dummy design. A double-dummy design is difficult to apply in the context of emergency medicine because of extremely tight time frames and may also raise ethical issues, as it could result in delayed administration of the active drug and a prolongation of pain for the patient with trauma. Finally, considering that the literature reports that the intravenous route is generally associated with a greater placebo effect, and that the medications which constitute SAT have widely been shown to be effective options for the management of moderate-to-severe trauma pain, the risk of bias in patient assessment is believed to be very limited in this specific case. The choice of SAT for moderate pain (IV ketoprofen or paracetamol) was discretional; however, most patients (88%) received paracetamol, thus providing a relatively consistent comparative treatment. We did not perform laboratory safety tests in this trial because of the setting and the lack of opportunity for in-person follow-up with patients; however, previous clinical studies have identified no renal or liver concerns arising from clinical laboratory evaluations performed at baseline and follow-up (14 ± 2 days) [
18,
32].
Acknowledgements
We thank the Contract Research Organization YGHEA, Division of Ecol Studio Bioikos S.r.l., Bologna, for the operational support and in particular Eleonora Romagnoli for protocol writing, Vanessa Bacchi for trial start-up, Elisa Lani and Francesca Bressan for clinical monitoring, Carolina Cerrè for data management and Francesca Cannoletta for project management. Penthrox® is a registered trade mark of Medical Developments International (MDI) Limited and used under licence.
List of Investigators
The investigators of the MEDITA Study Group are Germana Ruggiano, Guido Tota, Michela Zerini, Alessandra Semino, Sara Rasla, Elena Mecatti, Gabriele Panci, Catia Fronduti, Patrizia Signoroni, Alessandro Corradini, Alessandra Benucci (Emergency Medicine Department, Santa Maria Annunziata Hospital, Florence, Italy); Isabella Bartoli, Davide Messina (SUES 118 Catania-Ragusa-Siracusa, Cannizzaro Hospital, Catania, Italy); Giuseppe Carpinteri, Paola Noto, Chiara Giraffa (Department of Emergency Medicine, Vittorio Emanuele University Hospital, Catania, Italy); Andrea Fabbri (Department of Emergency Medicine, Morgagni-Pierantoni Hospital, Forlì, Italy); Francesco Bermano; Cristina Lapini (118 Service, San Martino Hospital, Genova. Italy); Maurizio Chiesa; Beatrice Sabini (Emergency Department, S. Antonio Hospital—ULSS 16, Padova, Italy); Mario Oppes, Vincenzo Pretti (Emergency Department, Sassari University Hospital, Sassari, Italy); Peppino Masciari, Antonietta Comito, Lucia Orlando (Emergency Department, Pugliese-Ciaccio Hospital, Catanzaro, Italy); Davide Torti, Giorgio Carbone, Gianfrancesco Alberto (Emergency Department, Gradenigo Hospital, Torino, Italy); Vittorio Iorno (Anesthesia, Resuscitation and Emergency Department, Fondazione IRCCS Cà Granda Ospedale Maggiore Policlinico di Milano, Milan, Italy); Antonio Voza, Alessandro Barraco, Fabio Intelligente, Monica Suardi, Claudio Addari (Emergency Department, IRCCS Humanitas Research Teaching Hospital, Rozzano, Milan, Italy); Piero Paolini, Nicola Bertocci (118 Unit, Azienda USL Toscana Centro, Pistoia, Italy); Sossio Serra, Stefano Tranelli, Pasquale Di Conza, Alessandro Valentino, Patrizia Cuppini, Elisabetta Lucchi, Carlo Garaffoni (Emergency Department, Maurizio Bufalini Hospital, Cesena, Italy); Gianfilippo Gangitano, Giuseppe D’Antuono, Lorena Bertozzi, Giulia Montanari (Emergency Department, Infermi Hospital, Rimini, Italy), Raffaella Francesconi (Faenza Hospital, Faenza, Italy).