Efficacy and safety of methoxyflurane: managing trauma associated pain in UK SAR helicopter paramedic practice

Efficacy

The search yielded five publications relating to the efficacy of MOF as an analgesic (Table 3). The highest quality of evidence was as a randomised, double-blinded, multi-centre clinical trial conducted by Coffey et al (2014) on patients aged over 12 years. Coffey et al (2016) later published the results of the adult subgroup from this study; it is this later publication that is included in this review.

This trial aimed to evaluate the short-term efficacy of MOF for the treatment of acute pain in adults presenting to an emergency department with a numerical pain score (NPS) of ≥4 to ≤7 as a result of traumatic injury. 203 adults met the screening criteria and were randomised to receive either MOF (n=102) via the Penthrox inhaler, or a placebo (PBO) (n=101). Study groups were similar in mean age, sex, race, presenting injury and initial pain score. A comprehensive description is given on how researchers ensured randomisation and blinding of both patient and attending staff. This included ensuring that PBO and MOF Penthrox inhalers were the same weight and smell. All patients were treated the same and all were accounted for throughout.

Pain intensity was recorded using the 100 mm visual analogue scale (VAS) before first inhalation, and at defined time intervals thereafter. Gausche-Hill et al (2013) conducted a systematic review to determine a preferred pre-hospital pain assessment tool. They concluded that there is insufficient evidence to support or refute a specific tool – the VAS is considered as valid and reliable a measure as any. These comprehensive, robust design elements ensure significant validity to the study (Coughlan et al, 2007; Hoare and Hoe, 2013).

MOF showed a highly significant greater analgesic treatment effect when compared to the PBO. The mean difference in the amount of pain reduction was −17.4 mm (CI −22.3 to −12.55mm, p<0.0001). In the MOF group, 82.4% (n=84) achieved analgesia compared to just 52.5% (n=53) taking the PBO. The time to first pain relief was significantly shorter with MOF, with 36.9% (n=37) subjects reporting relief within the first four inhalations compared to 11.9% (n=12) given PBO. 15.7% (n=16) given MOF gained no relief compared to 46.5% (n=47) given PBO.

Carley and Body (2014) expressed concerns over the ethical issues surrounding the use of a PBO in this trial. Coffey (2016) details the ethical approval and addressed the controversy of a placebo in this context by ensuring all participants gave consent and rescue analgesia was available. Within the first 20 minutes, 22.8% (n=23) of patients in the PBO group took the rescue analgesia compared to just 2.0% (n=2) in the MOF group.

This study is of high quality and demonstrates the efficacy of MOF. However, the lack of a comparator such as Entonox of IV morphine does detract slightly from its significance, but it is acknowledged that the use of such a comparator would make blinding impracticable.

Two systematic reviews of literature are included in this review, one focussing on MOF in trauma (Blair and Frampton, 2016) and the other evaluating the efficacy and safety of MOF in the pre-hospital and emergency setting (Grindlay and Babl, 2009). Both addressed a clearly focused question relevant to this research, detailing systematic methods for their evidence searches. In addition, they increased their reliability by not restricting themselves to published and/or electronic sources (Aveyard, 2014). All the evidence appears appropriate, although Blair and Frampton (2016) have omitted two seemingly relevant papers: Buntine et al, (2007) and Grindlay and Babl (2009).

Both reviews suffered from a paucity in high quality of evidence. Apart from the study conducted by Coffey et al, (2016) included in the Blair and Frampton (2016) review, the remaining evidence consisted largely of case studies and informative pieces. This moderates the reviews' rankings in typical evidence hierarchies (Aveyard, 2014; Cronin, Ryan and Coughlan, 2008).

While Grindlay and Babl (2009) offered justified assessments of the strengths and weaknesses of the evidence they reported, Blair and Frampton's (2016) review is largely descriptive. It presents as a factual précis of results, lacking in sufficient discourse and critique to satisfy the CASP's (2013) review checklist, thus diminishing the ability to attribute the otherwise deserved strength to its conclusions (Aveyard, 2014; Cronin, Ryan and Coughlan, 2008). However, the reviews' limitations do not significantly detract from their collaborative strength. Both reviews provide sufficient evidence of adequate quality that unanimously support the conclusions of Coffey et al, (2016) that MOF is an efficacious analgesic.

This is further reinforced by Buntine et al (2007). Their prospective, non-randomised, non-blinded observational case series aimed to determine the efficacy and adverse event profile of MOF. Adults presenting to Box Hill Hospital, Melbourne, who were administered MOF by ambulance crews were eligible to be enrolled. A NPS was recorded prior to MOF administration, then at pre-defined intervals thereafter.

The study was conducted over a 10-month period in a hospital that has around 40 000 attendances per annum, but only 83 subjects who met an appropriate inclusion criterion were enrolled. This was attributed to severely stretched staffing resources being unable to identify all eligible patients. Although the majority of the subjects (41%, n=34) were suffering pain as a result of traumatic injury, the study offers no other indication of its small convenience sample's representability. Hence, caution should be taken when generalising these results (Hoare and Hoe, 2013; Hunt and Lathlean, 2015).

Results were appropriately analysed using multi-variate statistical analysis to adjust for age, gender and injury type. Paired t-tests aptly measured significance (Freeman and Walters, 2015). The study had ethical approval and captured data using a questionnaire comprising of suitably chosen closed questions to record the presenting injury, NPS, observed side-effects and rate overall satisfaction (Oppenheim, 2000).

81.9% (n=68) of paramedics and 72.3% (n=60) of patients recorded that they were ‘satisfied’ or ‘very satisfied’ with the analgesic properties of MOF. Patients suffering pain as a result of acute musculoskeletal injury (n=34) gained a mean NPS reduction of 2.32 (± 0.37, p<0.0003), 0–5mins following self-administration of MOF. However, consideration must be given to the study samples' questionable representativeness and a lack of control placebo or comparator.

Middleton et al (2010) yielded more convincing MOF efficacy results in their retrospective, descriptive study that compared MOF alongside other analgesics. 42 844 adults presenting to the Ambulance service of New South Wales, suffering a NPS of ≥5 during a 23-month period were eligible for enrolment, the majority of whom were victims of traumatic injury. There were more than 102 455 potential subjects, but 59 611 (58%) had to be excluded based on incomplete recording of pain scores. Although again, this renders the results vulnerable to potential uncertain validity (Meadows, 2003), unlike the Buntine et al (2007) study, this is mitigated by the significant numbers in the MOF sample.

A total of 19 235 patients, 33% (n=6 656) of whom were victims of traumatic injury, self-administered MOF and experienced a statistically and clinically significant mean NPS reduction of 3.2 (3.1–3.2, p<0.001). Despite being inferior to morphine and fentanyl, MOF reduced NPS by ≥30% in approximately 60% of patients. These results appear to provide credible evidence to corroborate the unanimous conclusions of the other publications. MOF is an efficacious analgesic.

Patient safety

Patient safety was addressed by seven authors; a tabulated precis of the results is included in Table 4. Oxer (2016) investigated the deleterious effects that analgesic dosages of MOF had on cardiovascular and respiratory function. The sample was limited to patients who had been administered MOF between 15 August 2011 and 4 April 2012 by the St John Ambulance Service in Western Australia. It was then further limited to those who had at least three sets of observations recorded. This yielded a sample size of 590. Although no information was given on how representative this was of the total population, limiting data inclusion to trends rather than single readings boosts the measurements' reliability and validity. The majority of patients were 18 years of age or over (n=537, 91.1%) and 51.2% (n=302) were victims of traumatic injury, thus enhancing the results' applicability to the context of this research question (Coughlan et al, 2007; Hoare and Hoe, 2013).

Measurements of systolic blood pressure, pulse rate and respiratory rate were taken at defined intervals between 0 and 30 minutes following administration of MOF. The results were analysed using descriptive statistics, fittingly displayed numerically, complemented by graphs where appropriate (Freeman and Walters, 2015). Mean systolic blood pressure fell from 132.2 mmHg (SD 23.9) to 130 mmHg (SD 23.7). Pulse rate fell from a mean of 85.1 (SD 16.8) to 79.6 (SD 13.5) beats per minute. Mean respiratory rate fell from 20.4 (SD 4.9) to 18.6 (SD 3.9) breaths per minute. From an adverse side-effect inducing perspective, these reductions are insignificant.

This is reinforced by Coffey et al (2016). In addition to measuring pain intensity, Coffey et al (2016) recorded blood pressure, heart rate, respiratory rates and the presence of any adverse events (AE) at defined intervals up to 30 minutes post drug (MOF of PBO) administration. Patients also underwent a 14 ± 2-day follow-up where any subsequent AEs were recorded. To mitigate the subjectivity of AE observation, they were defined using a national toxicity criterion. Blood samples were also taken within −10 to +5 minutes of the start treatment, then again at the follow-up. Comparisons were made between liver and kidney function indicators.

In the Coffey et al (2016) study, there were no observable effects on cardiovascular or respiratory parameters following the administration of MOF. Mean changes in heart rate were ±5 beats per minute, blood pressure ±6mmHg and respiratory rate remained unchanged.

In the MOF group, 42.2% (n=43) of patients reported treatment related AEs compared to 14.9% (n=15) in the PBO group (Coffey et al, 2016). The most common AE was dizziness and headache, but these were more prevalent in the PBO group. All other AEs were reported by <5% of patients in either group. In the MOF group, AEs were mild and transient. Only one patient had to cease treatment due to vomiting, but ironically, they were from the PBO group. Buntine et al (2007) observed similar AE results in their study. Fifteen patients (18.1%) experienced minor MOF associated AEs. The most common were nausea (n=7, 8.4%), euphoria (n=3, 3.6%) and dizziness (n=2, 2.4%). There was one patient who became significantly sedated but recovered promptly when MOF administration was ceased.

The legacy of MOF is often blighted by the nephrotoxic stigma that led to its withdrawal from anaesthetic use. However, blood test results in these studies showed no evidence of nephrotoxicity or hepatotoxicity (Coffey et al, 2016). Although a valuable result, examining blood samples does not offer a conclusive indication (Dayan, 2016).

This potential weakness is addressed in a paper by Dayan (2016) who aimed to evaluate the potential nephrotoxicity of inhaled MOF in analgesic dosages. Although the methods' section describes a comprehensive search, there is no inclusion or exclusion criteria detailed. This, coupled with a relatively broad review question/aim, makes it difficult to establish if all the relevant papers have been included (Aveyard, 2014; Cronin et al, 2008).

From the evidence cited, a comprehensive, evidenced-based explanation is provided as to why the measurement of blood fluoride levels (a metabolite of MOF) appears to provide an appropriate indication of the deleterious effects of MOF. The review concludes that a level of <40µmol/L is considered safe. This corresponds to a ≤2 MAC hours, where MAC is the minimum alveolar concentration to produce surgical anaesthesia in 50% of healthy patients. This safe MAC limit is corroborated by numerous studies in the review.

The maximum dose of two vials of MOF via the Penthrox inhaler equates to 0.6 MAC hours, well below the derived 2.0 limit. Although a predominantly descriptive paper, Dayan (2016) convincingly concludes that the overwhelming majority of formal and clinical studies indicate that analgesic dosages of MOF are not associated with evidence of renal dysfunction. Blair and Frampton (2016) and Grindlay and Franz (2009) support this argument. Both successfully agree that analgesic dosages of MOF do not predispose patients to a greater risk of nephrotoxicity – this only occurs at higher, anaesthetic dosages.

In a study by Jacobs (2010), 17,627 patients were identified using an electronic record search as having received MOF via the Penthrox inhaler. This was from a total of 135,770 patients transported by the Western Australia Ambulance Service ambulance between 1990 and 2000. Trauma was the primary indication for use of MOF (n=9,755, 55.3%).

Occurrences of death, any ischemic heart disease, diabetes, cancer, hepatic or renal diseases were recorded for all patients and a comparison made between those administered MOF and those who had not. The follow-up period ranged from a minimum of four years to a maximum of 14 years after administration of MOF. This is far greater than any follow-up in other studies, significantly strengthening the reliability of the results (Hoare and Hoe, 2013). Jacobs (2010) includes comprehensive methods and analysis sections, which supports the validity of the results and their subsequent analysis (Hoare and Hoe, 2013).

In this study, patients who had received MOF in analgesic dosages via the Penthrox inhaler showed no greater risk of suffering the diseases investigated when compared to a similar group who had not received the drug.

Occupational safety

Around 19%–35% of MOF is exhaled unchanged which has occupational implications, particularly in confined spaces (Frangos et al, 2016). There were three pieces reporting on occupational safety which are listed in Table 5.

A narrative article by McLennan (2007) mentions anecdotal accounts of Australian Ambulance Service personnel experiencing nausea, headaches, vomiting and skin irritation when patients used analgesic doses of MOF in their vehicles. The reference for this discussion was a non-functioning internet link to an electronic press release LaborNET (2001, cited in McLennan, 2007). Although discussion pieces and expert opinion often rank at the bottom of evidence hierarchies (Hoe and Hoare, 2012; Aveyard, 2014), this is not sufficient to refute McLennan's (2007) plausible concerns. The implications of SARH crews suffering even the slightest of MOF induced occupational side effects at a critical stage in flight could be catastrophic. Prior to advocating the administration of Penthrox in flight, consideration must be given to this unlikely, yet potentially disastrous consequence.

There was no other evidence to support or refute this phenomenon in the search results, therefore the possibility of its occurrence was put to Galen Limited who market and distribute Penthrox in the UK and Ireland. Dr Sarah Dolan, Medical Manager for Galen Ltd, replied via email. In the UK and Ireland, the Penthrox inhaler has been modified to include an activated charcoal (AC) filter. Unpublished trials revealed that when the patient exhales into the Penthrox inhaler, the exhaled vapour passes through the AC chamber and exhaled MOF is absorbed. Dr Dolan stated that when tested at 40% MOF exhalation (above the 19%–35% typically exhaled), the percentage MOF concentration detected after passing through the AC chamber was ≤0.01% (negligible). This filter was not fitted to the inhalers used in the Australian cases cited by McLennan (2007), but its use is mandatory in the UK and Ireland to satisfy licensing conditions.

Frangos et al (2016) aimed to derive a safe maximum exposure level (MEL) and then compare this to the occupational levels typically experienced in an ambulance compartment or aircraft fuselage (Frangos, Mikkonen and Down, 2016). MEL derivation was based on a review of numerous MOF dose-response data studies. Although the evidence cited appears of significant quality, there is an apparent lack of systematism to its inclusion. There is no methods section and no inclusion or exclusion criteria listed; rendering it vulnerable to accusations of selection bias (Coughlan, Cronin and Ryan, 2007).

From the evidence cited, Frangos, Mikkonen and Down (2016) appropriately conclude that nephrotoxicity was the most relevant critical effect on which to base the MEL. The MEL was then derived from the data using the benchmark dose method rather than a no-observed-adverse-effect-level (NOAEL). This involved fitting an appropriate mathematical model to the numerous dose-response data sets. Unlike the NOAEL method, this enables the incorporation of more biological information into the process, thus boosting its robustness and validity (Filipsson et al, 2003).

In order to apply the MEL to the context of confined work spaces, data from two studies researching ambulance crews' exposure to MOF in Australia were reviewed. The studies provided robust data on breathing zone samples taken from ambulances whilst MOF was being administered. Workspace exposure (vapour) modelling, using a two-phase, steady-state-logarithmic decay model according to standard volatilisation principles was then applied.

The model assumed that health carers were being exposed to the patient inhalation of one vial per hour of MOF, over an eight-hour period via a Penthrox inhaler fitted with an AC chamber. Ventilation and workspace was assumed to be that of a standard ambulance compartment. It was also assumed that the total amount of MOF present in the compartment's air was being inhaled by a single health carer. Specific values relating to helicopter fuselages were not considered. However, these results appear reasonable, externally valid and applicable to this research question as the remaining conservative assumptions far exceed the realities of using Penthrox during SARH missions.

The model study concludes that the administration of MOF in confined workspaces would have no long-term health effects on those administering it. The MEL was 10 times higher than the amount calculated using the model parameters. Grindlay and Babl (2009) support this argument by stating that there were no causal links documented between MOF and any long-term occupational adverse effects.