Pre-hospital intranasal analgesia for children suffering pain: a rapid evidence review

Introduction: Pre-hospital analgesic treatment of injured children is suboptimal, with very few children in pain receiving analgesia. Studies have identified a number of barriers to pre-hospital pain management in children which include the route of analgesia administration. The aim of this review is to critically evaluate the pre-hospital literature, exploring the safety and efficacy of intranasal (IN) analgesics for children suffering pain. Methods: We performed a rapid evidence review, searching from inception to 17 December 2018, CINAHL, MEDLINE and Google Scholar. We included studies of children < 18 years suffering pain who were administered any IN analgesic in the pre-hospital setting. Our outcomes were effective pain management, defined as a pain score reduction of ≥ 2 out of 10, safety and rates of analgesia administration. Screening and risk of bias assessments were performed in duplicate. We performed a narrative synthesis. Results: From 310 articles screened, 23 received a full-text review resulting in 10 articles included. No interventional studies were found. Most papers reported on the use of intranasal fentanyl (INF) (n = 8) with one reporting IN ketamine and the other IN S-ketamine. Narrative synthesis showed that INF appeared safe and effective at reducing pain; however, its ability to increase analgesia administration rates was unclear. The effectiveness, safety and ability of IN ketamine and S-ketamine to increase analgesia administration rates were unclear. There was no evidence for IN diamorphine for children in this setting. Conclusion: Interventional studies are needed to determine with a higher confidence the effectiveness and safety of IN analgesics (fentanyl, ketamine, S-ketamine, diamorphine) for children in the pre-hospital setting.

are literature reviews that use methods to accelerate or streamline the traditional systematic review process (Ganann, Ciliska, & Thomas, 2010). As such, they are typically completed in compressed timeframes when compared to a systematic review.

Inclusion and exclusion criteria
In order to identify relevant studies that address the research question, the PICOS (participants, intervention, comparator, outcomes, studies) acronym was used (Table 1).

Participants
The search was restricted to paediatric patients, but included adolescents, so an upper age cut-off of under 18 years of age was chosen (United Nations, 1989). There was no lower age limit. Since it is possible that there are differences in the management of paediatric pain between the pre-hospital and in-hospital setting, both in terms of the range of available analgesics and the personnel who are likely to be undertaking that administration, only pre-hospital studies were included.

Intervention
The IN route of administration (in theory) removes some of the barriers to administering analgesics to children (Murphy et al., 2014;Williams et al., 2012). Specifically, this route does not require the infant or child to be old enough to understand the administration, which is required for inhaled analgesics such as Entonox ® for example. In addition, it is not as invasive as the intravenous route, which can be difficult to achieve and can cause further distress and pain.

Comparator
Where studies included comparisons with drugs administered via routes other than the IN route, the review attempted to compare the reported results that relate to the outcomes specified in this review. Non-pharmacological interventions, such as distraction and presence of parents, were not included as a comparator due to the low level of documentation and subsequent lack of representation within the literature, often being reported as a limitation (Browne et al., 2016;Jennings, Lord, & Smith, 2015;Lord, Jennings, & Smith, 2017).

Effectiveness
Effective pain reduction was defined as a reduction in pain score of ≥ 2 out of 10 using the numeric pain rating scale, Wong and Baker FACES ® scale or Face, Legs, Activity, Crying and Consolability (FLACC) scale. This measure Background According to McCaffery (1968, p. 95) 'pain is whatever the experiencing person says it is, existing whenever he [sic] says it does'. The World Health Organization (2017) and Lohman, Schleifer and Amon (2010) mandate that countries must provide pain treatment medication as a core obligation under the right to health. Pain can have psychological, physical and social consequences which impact on quality of life (Lohman et al., 2010). Without effective pain treatment, children are at risk of developing post-traumatic stress disorder (Saxe et al., 2001;Sheridan et al., 2014).
Studies have identified a number of barriers to pre-hospital pain management in children (Murphy et al., 2014;Williams, Rindal, Cushman, & Shah, 2012), which include route of analgesia administration, with the intranasal (IN) route proposed to overcome the challenge of cannulation in children. The aim of this review is to critically evaluate the pre-hospital literature exploring the safety and efficacy of IN analgesics for children suffering pain.

Research question
This rapid evidence review (RER) aims to inform a paediatric pain research working group on the utilisation of the IN route to administer analgesia to children in the pre-hospital environment. The following research question was proposed: Are IN analgesics safe and effective at reducing pain in children within the pre-hospital environment?

Objectives
The objectives of this RER are to: • search and evaluate the literature relating to the management of paediatric pain in the pre-hospital environment using IN and other routes of administration; and • present the findings to inform a paediatric pain research working group.

Safety
Safety of IN analgesics (incidence of adverse events and serious incidents).

Administration rate
Where reported, overall rates of analgesic administration will be assessed.

Studies
Randomised controlled trials (RCTs), quasi-RCTs and prospective and retrospective observational studies were eligible for inclusion. There was no restriction on language, but results were limited to research on humans. Editorials, position statements, letters, literature reviews, case reports and consensus statements were not eligible, but the references cited in these publications were reviewed and relevant papers included. Literature reviews were excluded to maintain a higher threshold of study quality. Qualitative studies were excluded due time constraints, considering the complex nature of meta-synthesis and meta-integration.

Study selection
A first pass of the search results was conducted independently by both authors, who screened the title and abstract against the inclusion/exclusion criteria to determine whether the papers might be suitable for inclusion. Once this was completed, the authors came together to address any disagreements about paper suitability. An independent arbiter was available for disagreements relating to inclusion that could not be resolved.
The full text of papers that made it through the first-pass process was obtained and independently reviewed (second pass) by both authors. Those that met the inclusion criteria were put forward for inclusion in the review. Fulltext papers which were excluded at the second-pass stage were still screened for potentially relevant studies to inform the review.

Critical appraisal
Papers that successfully made it past the second-pass process were critically appraised. The robustness of the evidence was determined by evaluating each paper against a list of criteria described by Collins et al. (2015). Separate lists of criteria were available for quantitative interventional and observational studies. Each criterion was given a score of 1-3 (1 being the lowest), and from these an overall critical appraisal score of 1-3 was awarded, based on the most commonly awarded score for each criterion. However, studies were not excluded from the review as a result of their risk of bias assessment given the anticipated low numbers of studies available. See Supplementary 2 for the risk of bias assessments.

Data synthesis
Once the final papers for inclusion had been selected, a synthesis of the evidence was conducted. This consisted of: 1. describing the volume and characteristics of the evidence base; 2. utilising the synthesis to answer the research questions; 3. highlighting the implications of the findings; and 4. making recommendations for further research. of bias', with Murphy et al. (2017) and O'Donnell et al. (2013) being highly relevant and Karlsen et al. (2014) and Pasquier et al. (2017) less relevant to the target population. Johansson et al. (2013) and Reid, Hatton and Middleton (2011) were deemed at 'high risk of bias', with poor relevance to the target population and outcome measure.

Effectiveness
Intranasal fentanyl Murphy et al. (2017) reported on the use of intranasal fentanyl (INF) in children < 16 years (n = 94) and found that a clinically effective reduction in pain score occurred in 78 children (83%; 95% confidence interval (CI) 74-89%). Lord et al. (2017) reported on the implementation of INF in children < 15 years (n = 9833) and found that before the intervention, 88.1% (n = 3114) of children receiving analgesia had a reduction of pain severity of 2 or more points, with 94.2% (n = 5933) achieving this benchmark after intervention (p < 0.0001). The odds of a reduction in pain of 2 or more points increased by 1.01 per month immediately before the intervention and 2.33 after intervention (p < 0.0001). Lord et al. (2016) reported on the use of INF in children < 15 years (n = 38,167) and found that for those receiving fentanyl (95.6% of the study population received this via the IN route), the initial median pain score was 8 (interquartile range (IQR) 6-9) with a median pain score change of 5 (IQR 3-7).

Results
A total of 310 articles received a title and abstract screen; 23 of these received a full-text screen, which resulted in 10 articles included in this review (Figure 1). A summary of included studies can be found in Table 2.
Ten papers were included in this review and none were high-quality interventional trials. There were eight observational studies, one case series and one case report that met the inclusion criteria. These papers reported on the use of three IN analgesics: fentanyl (n = 8), S-ketamine (n = 1) and ketamine (n = 1). Seven studies reported on the effectiveness, four on the safety and four on the administration rate of IN analgesics administered to children (< 18 years) in the pre-hospital setting. Papers were deemed to have a 'low risk of bias' (n = 3), 'moderate risk of bias' (n = 5) and 'high risk of bias' (n = 2). The relevance of these studies to the research question, target population and outcome measure was deemed 'high' (n = 6) and 'low' (n = 4). Bendall, Simpson and Middleton (2011) and Lord, Jennings and Smith (2016   Median reduction in pain score before and 5 to 10 minutes after administration of nasal S-ketamine. Effectiveness: VAS-score decreased from a median of 10 (IQR 7.5-10) to 3 (IQR 2-4), p = 0.0018. For children < 18 years (n = 6) decrease from a median of 10 (IQR 8-10) to 2 (IQR 1-3). Safety: side effects in these nine cases were few and non-serious. Administration rate: n/a. 'Effective analgesia' defined as a reduction in pain score of ≥ 30% using the VNRS-11. Effectiveness: all analgesic agents were effective in the majority of patients (87.5%, 89.5% and 78.3% for morphine, fentanyl and methoxyflurane, respectively). For those receiving INF alone, initial median (IQR) pain was 8 (7-10) with a median (IQR) pain score difference of 5 (3-7). There was evidence that methoxyflurane was less effective than both morphine (OR 0.52; 95% CI 0.36-0.74) and fentanyl (OR 0.43; 95% CI 0.29-0.62; p < 0.0001).

Risk of bias and relevance to study question
There was no clinical or statistical evidence of difference in the effectiveness of fentanyl and morphine in this population (OR 1.22; 95% CI 0.74-2.01).
There was no evidence that combination analgesia was better than either fentanyl or morphine alone.
Safety: n/a. Administration rate: n/a.  Karlsen et al. (2014) reported on the use of INF in adults (n = 840) and children over eight years (n = 63) and found that in those aged < 18 years, the median reduction in pain score was 4 (IQR 2-5), with 87% achieving clinically relevant reductions (defined as a pain score reduction of ≥ 2 points out of 10). Bendall et al. (2011) reported on the use of INF in children aged 5-15 years (n = 3312) and found that 89.5% of patients who received INF achieved effective pain reduction (defined as a reduction in pain score of ≥ 30% using the 11-point verbal numeric rating scale, VNRS-11). For those receiving INF alone, the initial median pain score was 8 (IQR 7-10) with a median pain score difference of 5 (IQR 3-7). There was evidence that methoxyflurane was less effective than fentanyl (odds ratio (OR) 0.43; 95% CI 0.29-0.62; p < 0.0001), but no clinical or statistical evidence of difference in the effectiveness of fentanyl and morphine in this population (OR 1.22; 95% CI 0.74-2.01). There was no evidence that a combination of analgesics was better than either fentanyl or morphine alone.

Intranasal ketamine/S-ketamine
Ketamine, or racemic ketamine, is a mixture of optical isomers. S-ketamine is a purer, single isomer of ketamine with twice the anaesthetic and analgesic potency of racemic ketamine (Doenicke, Kugler, Mayer, Angster, & Hoffmann, 1992). Johansson et al. (2013) reported on the use of IN S-ketamine in adults (n = 3) and children (n = 6) in an alpine setting, and reported that for those aged < 18 years (n = 6) median pain scores reduced from 10 (IQR 8-10) to 2 (IQR 1-3). Reid et al. (2011) reported on the use of IN ketamine in one paediatric patient suffering a thermal injury and found that a satisfactory level of analgesia and anxiolysis (reduction of anxiety) was achieved, allowing the comfortable application of a burns dressing and patient transport. Bendall et al. (2011), Karlsen et al. (2014, Lord et al. (2016Lord et al. ( , 2017 and Murphy et al. (2017) all reported high levels of effective pain management when using INF, suggesting that INF is effective at reducing pain in children within the pre-hospital setting. Johansson et al. (2013) and Reid et al. (2011) also reported effective pain management via the IN route, although with S-ketamine and ketamine, respectively. However, the high risk of bias and low relevance with respect to this review indicates that further research is required to assess the effectiveness of IN S-ketamine/ketamine for the treatment of pain in children in the pre-hospital setting.

Safety
Intranasal fentanyl Pasquier et al. (2017) reported on the use of INF in adults (n = 912) and children (n = 244) and found no serious adverse events, although, as a retrospective review, it has to be acknowledged that side effects could have been under-detected and/or under-reported. The authors concluded that fentanyl is a safe analgesic, with minimal cardio-respiratory repercussions. Karlsen et al. (2014) reported 39 potential adverse events in 36 patients, none of which were serious. However, it was not possible to determine the proportion of adverse events that occurred in children in this study. Finally, Murphy et al. (2017) recorded no adverse events during their study (n = 94). Johansson et al. (2013) did not explicitly quantify all side effects, although they did confirm that no patients had airway compromise. Three did experience vertigo (not specified whether adult or child) and 'many' complained of the taste.

Administration rate
Intranasal fentanyl Pasquier et al. (2017) found that the proportion of paediatric patients receiving analgesia was significantly higher when INF was available in comparison with the intravenous route only (57% vs. 22%, p < 0.001). This difference also remained statistically significant after limiting the analyses to only the 71 cases where INF was used as the first analgesic route and drug. Browne et al. (2016) reported on the use of INF in children < 18 years (n = 7340) before and after treatment protocol amendments aimed at improving paediatric pain management. They found a reduction in INF administration rates from 27% (n = 45) before the protocol change, to 17% (n = 32) after (p = 0.02). In addition, there was no increase in opiate administration to eligible patients across study sites, regardless of documented pain severity. This remained at 5% throughout the study period.
The IN route was the most common as reported by Lord et al. (2016), with 95.6% of children < 15 years of age who were administered fentanyl receiving it via the IN route. In contrast, O'Donnell et al. (2013) reported on the use of fentanyl in children < 16 years (n = 233) before and after the introduction of a mucosal atomisation device (MAD) and found no statistically significant difference in the rate of fentanyl administration between the pre-MAD (30.4%) and post-MAD (37.8%) groups.

Discussion
Effectiveness INF appears to be effective at reducing pain in children in the pre-hospital setting. The effectiveness of IN S-ketamine/ketamine could not be established given the low numbers of patients and poor quality of the included studies. There was no evidence of the effectiveness of other IN analgesics such as diamorphine for children in the pre-hospital setting. effectiveness, safety and ability of IN ketamine/S-ketamine to increase analgesia administration rates could not be determined. There was no evidence relating to the use of IN diamorphine for children in the pre-hospital setting. Further research is required, preferably utilising an interventional approach, to improve the quality of evidence.

Author contributions
Both authors made substantial contributions to the conception and design of this review; the acquisition, analysis and interpretation of data; drafting the work and revising it critically for important intellectual content; and final approval of the version to be published. They agree to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.

Conflict of interest
GAW declares no conflict of interest. RP is the editor-inchief of the British Paramedic Journal but had no editorial control over this publication.

Ethics
Not required.

Funding
None.

Safety
Overall, none of the included papers reported any serious adverse effects from analgesics administered via the IN route. INF appeared safe, but the poor relevance of Karlsen et al. (2014) and Pasquier et al. (2017) and the small sample size of Murphy et al. (2017) limit the strength of this finding. The safety of IN ketamine could not be established with any confidence, given the high risk of bias and poor relevance, compounded by a small sample size overall (n = 7) (Johansson et al., 2013;Reid et al., 2011).
Evidence for the safety of INF was limited, but suggested it was safe, with some minor but no serious side effects reported. The safety of IN S-ketamine could not be established with any confidence. There was no evidence to determine the safety of IN ketamine or diamorphine for children in the pre-hospital setting.

Administration rates
Evidence regarding the administration rates of INF were conflicting and therefore could not be established. However, the studies from the United States appear to be outliers in terms of lower than expected paramedic utilisation of IN routes for paediatric analgesia. There was no evidence to determine the administration rates of IN S-ketamine/ketamine or diamorphine for children in the pre-hospital setting.

Strengths and limitations
The lack of interventional studies limits the quality of data available for synthesis, meaning we were not able to determine causation. We were unable to determine whether INF causes effective pain management, so can only conclude that it appears to be effective.
The rapid nature of this review meant that we only searched a limited number of sources and did not search trial registries or perform forward/backward citation tracking; therefore, there is a possibility that some articles have been missed. We did not determine the confidence in the cumulative evidence.
We believe a strength of this review is the limitation to the pre-hospital context. This allows for simpler analysis and interpretation and the development of clear recommendations for future research in this setting.

Recommendations for future research
Interventional studies are needed to determine with a higher confidence the effectiveness and safety of IN analgesics (fentanyl, ketamine, S-ketamine, diamorphine) for children in the pre-hospital setting.

Conclusion
INF appeared effective and safe, but its ability to increase analgesia administration rates was unclear. The