Introduction

The cliché that states “children are just small adults” is certainly not true in the case of long bone fractures. A child’s experience of long bone fractures is dramatically different from that of an adult on account of their rapidly developing physiology (Wood et al 2003). This rapid development results in biochemical and physiological differences between a child’s and an adult’s skeleton, the mechanisms of fracture and healing, are an important component of their treatment needs and consequently crucial part of emergency care management (Bonadio et al 2001). In addition, children, from infancy through to adolescence, have common fracture patterns related to their stage of development. The structural differences between the bones of a child and an adult enable children’s bones to endure greater forces and to heal quicker – a child’s remodeling potential supports full recovery with limited or no long term side effects from long bone fractures (Lane et al 1998).

Injuries of all types are the second leading cause of hospitalization among children younger than 15 years (Landin 1997). Musculoskeletal trauma, although rarely fatal, accounts for 10% to 25% of all childhood injuries (McDonnell 1997, Landin 1997, Lane et al 1998). Boys have a 40% risk and girls a 25% risk of incurring a fracture before the age of 16 years (Landin 1997, Ritsema et al 2007). The most common site of fracture is the distal forearm which accounts for 50% of paediatric fractures. The rates of fracture increases with age as children grow; peaking in early adolescence.

Fortunately, most fractures in children are minor – greenstick and torus fractures constitute approximately 50% of all fractures in children (Landin 1997, Lane et al 1998, Gasc & Depalokos1999, Richards et al 2006) and only 20% require reduction. Thus, the management of paediatric fractures is often straightforward. Without exception children will experience pain at the time of injury, attending the accident and emergency department and during recovery. The most common pain management strategies involve a multi-modal approach that includes both pharmacological and non-pharmacological components delivered via the least invasive technique (Worlock et al 2000). In practice this includes oral medication, such as oramorph, paracetamol, and NSAIDs, inhaled entonox, intranasal diamorphine (IND) or intravenous opioid where necessary and distraction with age appropriate devices, such as interactive books, bubbles, music and computer games in older children. Notably, IND is currently embraced as the key route of opioid delivery for children attending AED with fracture pain in the UK British Association for Accident and Emergency Medicine (BAAM E 2002).

Parents and guardians of children frequently seek care in AED for the relief of pain from traumatic injuries and as a result the field of emergency medicine has assumed a leadership role in paediatric pain management. However, despite this the literature suggests the provision of pain relief for children attending AED remains suboptimal when compared to adults with the same injuries. Further discrepancies are reported between paediatric accident and emergency departments (PAED) and district general accident and emergency departments (DGAED) (Emergency Triage 2004). One reason suggested for these differences is the geographic distribution of specialised services, which are predominantly located in large cities where they are affiliated with universities. However, a recent audit by the British Association for Emergency Medicine (BAAEM 2005) of their guideline for the management of pain in children shows inconsistencies in provision of analgesia particularly for fracture pain throughout the country with no measurable difference between PAED and DGAED. A key feature of this guideline is the algorithm which advocates the use of IN diamorphine for acute moderate to severe pain in children over the age of one year (see appendix 1).

The whole topic of analgesia in the paediatric population is complex and still imperfect especially in acute moderate to severe pain requiring urgent treatment in the emergency department (Schechter et al 2002). The road to pain free suffering is still paved with impediments such as failure of pain recognition and methods of delivery of analgesia (Murat et al 2003). Oral administration can be inadequate in an emergency situation with particular limitations in potential choice of drug and delay in gastric absorption and gastric emptying. Intramuscular (IM) and intravenous (IV) administration can be distressing to children and have been shown to influence future response to painful procedures (Gidron et al 1995, McGrath et al 2000, Fitzgerald et al 2005, Walker et al 2007). Rectal administration has limited acceptability given unpredictability of onset together with occasional problems of consent (Mitchell et al. 1995). By contrast, the efficacy and safety of the IN route has been well documented for desmopression acetate (DDAVP), insulin, antihistamines, midazolam and calcitonin (Jewkes et al 2004, Loryman et al 2006). In contrast, intranasal administration has a number of advantages. It is technically straightforward, socially acceptable and demonstrably effective. The nasal mucosa is richly vascular and administration by this route avoids the first-pass metabolism phenomenon

Summary

Studies in the 1990’s such as Yearly & Ellis (1992) have also demonstrated the efficacy of administration of intranasal medication via a nasal spray rather than drops in adults, although the efficacy of this application in the paediatric population remains to be proven. Intranasal administration is possibly the ideal route of analgesic administration in children. Currently, within the accident and emergency department (AED) of Bristol Royal Hospital for Children (BRHC) intranasal diamorphine is used as the first rescue analgesia in the paediatric population presenting with acute moderate to severe pain, most frequently in patients with long bone fractures who do not require intravenous access for resuscitation.

Diamorphine is a semi synthetic derivative of morphine with a number of properties that render it a desirable analgesic agent for administration via the nasal route. It is a weak base with a pKa of 7.83 and is water soluble allowing high concentration to be administered in small volume (Rook et al 2006). Unfortunately the legal use of diamorphine is limited to two European countries i.e. United Kingdom (UK) and Sweden. Furthermore periodic problems with its availability during the past few years (with further shortfalls in availability predicted by the NHS purchasing and supply agency) have resulted in an alternative efficacious analgesia being sought for this population.

Fentanyl, however, is a short rapidly acting opiate has several qualities that render it useful as an IN analgesia and a potential candidate to replace IN diamorphine in the AED for acute facture pain management in children. It has a very high lipid solubility, potency and diffusion fraction, and unlike diamorphine it is not a prodrug and does not cause histamine release (Reynolds et al 1999).

Assessment of a patient’s pain experience is not directly accessible to others, collecting and analyzing information about the processes of pain relief and pain prevention is not straightforward and presents significant challenges to health care professionals. In children, this task is further complicated by their varied stages of physical and cognitive development.

Recent research by Bruce & Frank (2004) however, has shown that the ability to measure pain in the paediatric population has improved dramatically and that today there now exists a plethora of age appropriate pain assessment tools for acute pain in children ranging from pre-term infants to adolescents, the majority claiming validity (strength and robustness) and reliability (consistency). However, most clinical research into pain management strategies continue to rely on the gold standard – self report and visual analogy score tools (mostly 0-10) (Chalkiadis 2001, Walker et al 2007). Although these tools are reliable they are not always adapted appropriately for a child’s stage of development. Children’s understanding of pain and their ability to describe pain change with increasing age in a developmental pattern consistent with the characteristics of Piaget’s preoperational, concrete operational and formal operational stages in cognitive development (Smith et al 2003). The quality or intensity of the pain can be difficult to determine in children, as most tools rely upon a patient’s relative judgment between the intensity of present pain versus a patient’s worst pain experience (Murray et al 1996). These tools can therefore be unreliable where a child’s age of development means they have limited or no memory of pain experience.

Stevens et al (2002) recently described a conflict of understanding that resulted in a study bias and an insignificant reported power of (p=0.6). In the study an 8 year old boy had chosen the VAS (0-10) but frequently reported his score as 10, although he understood the increasing value of the scoring system further questioning identified he perceived 10 of 10 to be a good score and 0 of 10 to be poor. The boy was at a stage of development that limited his understanding of less is more. This case highlights the importance of utilizing a pain assessment technique that reliably accounts for a child’s age of development.

A preliminary search of literature suggests there is currently exists limited research to support for the use of intranasal diamorphine or intranasal fentanyl for the management of acute pain in long bone fracture in children as evidenced based medicine. Despite this lack of evidence it remains a key strategy within paediatric AED for the pain management of long bone fractures and is anecdotally reported as a gold standard for paediatric pain management. Therefore; its lack of availability could profoundly compromise pain management for this population. Thus, this extended literature review will examine the efficacy of intranasal fentanyl as an alternative to intranasal diamorphine for traumatic fracture pain in children attending accident and emergency departments. However, in these days of evidence based medicine, it clearly needs to be established beyond all reasonable doubt. In view of that only research into paediatrics will be included increasing the credibility of its application to practice.

SEARCH STRATEGY

A range of complimentary search techniques were used to capture key research including a systematic electronic literature search of the Cochrane library, Embase, CINAHL, Proquest, Medline, PubMed since 1990 up to 2009 (this has to be to year of submission). The scope of the search was extended beyond the recognised five years of current research so as to include the empirical work into the development of IN analgesia in children.

Key words used included the following: pain, acute pain management, intranasal diamorphine, intranasal fentanyl, procedural, accident and emergency, emergency department, child, pediatric, paediatric, child and fracture pain, as well as various combinations. In addition, in order to ensure the completeness of the search, an internet search was completed using the Google search engine, IASP, Pain Journal, Paediatric Nursing, BAAEM, NICE, Medline, EBM; the RCN was also utilised. Backward chaining of references found was also performed to ensure all relevant papers were identified.

Although this review identified twenty seven citations it should be noted that historically there are fewer Randomised Controlled Trials (RCT) in children compared to adults possibly due to problems gaining ethical approval and consent. Additionally even experienced researchers will be unable to find all relevant papers and much research is not submitted for publication. The studies identified were divided into the three modalities of IN route, IN diamorphine and IN fentanyl with the majority presenting evidence for the IN route.

All papers were critiqued using a tool published by the Learning and Development Department within the Public Health Resource Unit of the NHS (www.phru.nhs.uk/casp). The tool facilitated critiquing different forms of quantitative research and is based on work by Sackett (1986), Sackett et al (1996) and Phillips et al (2008) (see appendix 2). The results of the critique process for each paper and level of evidence applied in line with the modalities they address informed understanding of current practice and development of a research proposal.

STRUCTURE OF THE LITERATURE REVIEW

This literature review will focus on determining whether IN fentanyl is an effective alternative to IN diamorphine for the management of long bone fracture pain in children attending an AED. The scope of the literature review considers literature from 1990 onwards although occasionally earlier research has been referenced.

Given the limited available evidence on the topic the following review structure has been selected. Chapters 1, 2 & 3 will present the evidence sourced on each theme intranasal route, intranasal diamorphine and intranasal fentanyl with a short summary to conclude each chapter. Chapter 4 will present an in-depth discussion and conclusion on the utility of the evidence, its application to practice and the requirement for a multi-centred comparative randomised control trial to improve the credibility of the evidence base for this field of treatment. Finally chapter 5 will present a research proposal for a comparative study of these modalities.

Intranasal (IN) route of medication delivery in children.

Nasal administration of drugs has been reported as having several significant advantages over current practice which are predominately oral, IM, IV and rectal (Williams & Rowbotham 1998). It is emerging as a low-tech, inexpensive and non-invasive first line method for managing either pain or other medical problems (Wolf et al 2006). Nasal medication delivery takes a middle path between slow onset oral medications and invasive, highly skilled delivery of intravenous medications. The nose has a very rich vascular supply, IN facilitates direct absorption to the systemic blood supply due to increased bio-availability of the drug by missing first pass metabolism, It avoids the potentially technically difficult of sterile intravenous access, is essentially painless and is considered acceptable to children when compared to other routes of administration (Shelly & Paech 2006) (see table 1). a theory which will be considered when reviewing the studies within this chapter Therefore suggesting the IN route will result in therapeutic drug levels, effective treatment of seizures and pain without the need to give an injection or a pill, furthermore; it is quite inexpensive, an advantage in this era of increasingly expensive medical technology (Shelly & Paech 2006). Additionally given the complexity of the developing child and the known consequence of poorly managed pain on the future responses to pain the IN route does, if it is as efficacious and as safe as suggested offer one of the most acceptable, definitive forms of analgesia delivery in children.

The degree of accuracy of the previous statements will be established within this chapter by critically reviewing the 16 studies identified on IN medications other than intranasal diamorphine or intranasal fentanyl in the paediatric population (see table 2) as these agents are considered individually in later chapters. The rigour of the studies will be addressed within this chapter and reflect the level of evidence applied according to Sackett (1986) criteria (see appendix 3). Most studies reviewed were randomised clinical trials and in some cases compared against a placebo

Conversely, this does not concur with the trials discussed earlier (Lahat et al 1998, Al-rakaf et al 2001, Fisgin et al 2002, Mahmoudian and Zadeh 2004 and Holsti et al 2007) where significant dosing was applied or in Wilson et al (2004) who retrospectively studied 30 children age 2-16 years receiving 0.3mg/kg at 5mg/1ml INM and 13 patients receiving rectal 0.2mg/kg diazepam for seizures. The authors report equal efficacy for both routes. Success of these agents was considered on cessation of seizures, no reported complication and not needing to attend A+E. A total of 27/30 families who had used INM found it effective and easy to use. Although 20/24 (83%) who had previously used rectal diazepam still preferred it mostly due to the coughing and the volume of liquid administered via the IN route. Given it is generally considered that the optimum IN dose as stated above is 0.1- 0.2 ml per nostril, all but the studies discussed so far were using drug concentration and dosing regimes which resulted in large volumes of liquid being dripped in to the nasal cavity.

This is particularly poignant in Wilson et al (2003) who compared buccal to IN midazolam in 53 children aged 3-12 years experiencing seizures lasting > 5 minutes attending AED. A key feature of this study is the mean age of the children (age 9 years), mean weight (24kg) the study drug concentration as with previous studies was of 5mg /ml. IN dosing was at a dose of 0.3mg/kg. Given these figure the average dose would have been 7.2mg = a volume of 1.4ml being administered. Since the comparative route of administration for this study was buccal there is a possibility that part of the IN dose was buccally absorbed therefore creating a flaw in this study methodology, raising questions over why this comparative route was chosen and suggesting the only real conclusion to be taken from this particular study is buccal midazolam is effective and safe in children. Furthermore although this is described as a blind RCT and the authors claim the time to cessation of seizure was quicker for the INM group 2.43 (SD 1.67) to 3.52 (SD 2.14) for buccal route there is little detail on the blinding process or data collection procedure suggesting the rigour of the study maybe flawed therefore the efficacy and safety claimed for the IN route should not be embraced without further study.

On the other hand Fisgin et al (2002) and Hardord et al (2004) compared the INM with rectal diazepam. In Fisgin et al (2002) in an unblinded RCT equivalence study the authors compared INM with rectal Diazepam to ascertain the safety and efficacy of INM for the development of a clinical protocol in the management of prolonged seizure in children attending the AED. Forty five infants and children age 1 month -13years experiencing prolonged seizures > 10 minutes were either given INM 0.2mg/kg or rectal diazepam 0.3mg/kg. The authors report proven efficacy (p < 0.05) for INM stating time to cessation was faster INM 2-5 minute Vs rectal diazepam 5-7 minutes. Additionally, only 3 patients in the INM group compared with 10 in the rectal diazepam group needed a second dose. Statistical analysis applied to parametric and nonparametric values, respectively was SPSS Inc, The Mann-Whitney U-test, Wilcoxon rank sum test and Pearson chi-square test. Interestingly group stratification was almost identical, further supporting the study outcomes claimed. However it is worth noting that the sample size is small considering this is an equivalence study and without detail on the process of powering the study the possibility of a false positive result should not be ignored. As power for a study increases the chances of a false negative result decreases, therefore, appropriate powering is a vital component of clinical trials and should be preformed early in any study development

Intranasal Diamorphine (IND)

The delivery of opioids via the IN route is perhaps one of the most valuable indications for IN medication delivery. Acute pain is a frequent experience for children whether attending an AED, hospital and hospice setting (Hamer et al 1997). Furthermore it is not unusual for them to experience frequent episodes of breakthrough pain which requires additional support from fast acting analgesic agents. Owing to the developmental and physiological difference in the paediatric population there is a need for a variety of effective treatment option from which to select and individualise the patient’s therapy to meet their needs. IN opioid is simply one such option available which may be useful in children. It has been suggested that the delivery of medications via the IN route results in rapid absorption with medication levels within the cerebral spinal fluid (CSF) being comparable with (IV) administration (Chien and Chang 1997).

Diamorphine hydrochloride is a semi-synthetic derivative of morphine. It is extremely hydrophilic, which makes it ideal to use when preparing in high concentrations in solution, thus allowing high doses to be administered in smaller volumes via the intranasal route (Kendall & Latter 2003). However, this route of administration can be a painful process as reported by adults (Henry et al 1998). Despite this the intranasal route is considered more acceptable to children and their parents and is thought to lessen the opioid side effect profile seen in IV administration (Stoker et al 2008). This concept has been well recognised throughout the UK and many centres already use intranasal diamorphine for acute pain in children, following the guidelines by the British Association for Accident and Emergency Medicine Clinical Effectiveness Committee (2002) (BAAEM). Although the administration of intranasal diamorphine is now a first line choice for moderate to severe acute pain for children attending AED, as is the case within our institution, there is very limited research to substantiate this practice although as noted above it has been readily accepted by the BAAEM for acute pain management in children and very successfully used within our institution A recent shortage of diamorphine evoked the search for an equally effective and acceptable alternative.

Early research in animals and adults reported pharmacokinetics of nebulised inhalation and intranasal administration of diamorphine as detected morphine in plasma at six minutes (Masters et al 1988, Kendall 2001). Despite the age of this research and the fact that the later study was in adults, it is still quoted as creditable evidence to support this practice in paediatrics. However the legitimacy of this should be questioned, due to “children not being just small adults” but have physiological differences intrinsic to their age and stage of development which may affect the body’s absorption and level of toxicity in different ways to adults.

The extensive literature search highlighted four randomized controlled trials (RCT) that demonstrate IND to be clinically superior to intramuscular morphine and inferior to IV morphine particularly in the management of acute pain in children, a case study of an 8 year old boy and clinical audit of IND for pain relief in children attending AED (see table 3). The key methodology in the RCTs by Wilson et al (1997), Kendall et al (2001), Brennan et al (2004) and Brennan et al (2005) suggest these are superiority studies where the authors hypothesised improved pain management with the IND when compared to a variety of routes. The rigour of the studies will be discussed later in the chapter. Although while the critiquing process takes place it is fundamentally accepted that RCT are considered level 1 or 2 evidence as opposed to case study or audits at Level 3b and therefore generally sourced to

Latest published clinical evidence to support the use IND in the paediatric population is presented in an audit by Gahir & Ranson (2006) of 54 children whose care was managed by the use of an integrated care pathway for acute pain management while attending the local AED. This integrated care pathway focused strongly on the use of IND. Data collection was on a one page performa and included consent, date, patient demographic, pain score and side effect profile. Data collection was retrospective and data analysis illustrated limited recording of side effect profile but improved pain scores. However only 60% of patients have this information documented so data collection was difficult.

Despite this lack of hard evidence no clinical incident, including the side effect profiles, were reported. Thus suggesting the practice of IND for acute fracture pain management in children could be safe, effective and more acceptable to children than the more painful alternative of IM or IV administration. However there is limited strength in an audit, other than a review of practice (Bowling & Ebrahim 2005) and in this case a key feature for review should be the documentation process in the department as there were facets in the care pathway administration documentation missing. Therefore this audit suggests that IND is safe and effective pain management for children, but this conclusion can not be categorically drawn from the limited data available. The potential outcome of this audit could be education on documentation, to do a more rigours prospective audit of practice. Unfortunately at this point it only offers an insight to their clinical practice which is favourable for this agent and route. Albeit as noted before IND has improved children’s pain management and over all experience of acute care in our PAED additionally as with the results of the audit we have experienced no side effects or complications, further highlighting the importance of seeking an alternative to IND which offers equally efficacy.

Intranasal Fentanyl (INF)

Monitoring of the usual observations and pain scoring in the child was recorded prior to the administration of fentanyl (20 micrograms for 3-7 year olds and 40 micrograms for 8-16 yrs) and continued at 5 minute intervals for the 30 minute period. Additional doses of fentanyl (20 µg) were available if required at 5 minute intervals. Pain assessment was achieved with two validated pain assessment tools, the visual analogue scale (VAS) in older children and the Wong-Baker Faces (WBF) for younger children. Both are reliable and known to support consistency in pain assessment. Though there was no mention of training for those assessing this primary end point using these tools in the paper therefore this should be considered in the overview of the standard of evidence produced by this study. Additionally although forty five patients were randomized following consent unfortunately no details on the randomization process was disclosed in the paper either. This may not be significant, but when reviewing the credibility of the authors claims these obvious omissions could be responsible for a flaw in this study and remains to be established.

On the other hand, the methodology that has been disclosed in the paper appears sound as it addresses key areas of sample calculation (power of the study) as a superiority study with the sub groups size adequate to detect a significant difference (Greenhalgh 2004); demographics’, blinding of the drugs, assessors and appropriate statistical analysis of the data therefore supporting the validity of the results claimed and the application of the results to the age of patient targeted that this literature review is aiming to find an analgesic alternative to IND for.

The results concluded by Borland et al (2002), are a reduction in pain score at 10 minutes to 44.6 mm (95% confidence interval) 36.2-53.1 mm from 62.3 mm 53.2-69.4 mm (95% confidence interval) at assessment using the VAS and 2.2 (95% confidence interval 1.3-3.1) at 10 minutes from 4.0 (95% confidence interval 3.3-4.7) at assessment in 16 children using WBS. Visual analogue pain scores demonstrated clinically significant reductions in pain scores by 5 minutes that persisted throughout the entire study (up to 30 minutes) for both INF and IV morphine. The second primary end point of this study (side effect profile) showed no significant change in physiological parameter of the children’s pulse or respiratory rate, blood pressure or oxygen saturations, interestingly the side affect profile chosen for monitoring such as pulse and blood pressure are not considered to be one of the primary side affects of morphine, however nausea and vomiting which are was not assessed. Ultimately, there were no negative side-effects and the sizeable reduction in pain scores (compared to baseline assessments) was accomplished in children using INF by 10 minutes and maintained throughout the 30 minute period with the mean INF dose at 1.5µg/kg and ranging from 0.5-3.4 µg/kg. Interestingly 35.5% of children in the INF group only required one dose. Given the clinical equivalency of these two agents and routes the author’s conclusion that INF offers the benefits of a simple painless technique for treating acute pain is substantiated.

These benefits suggest that the IN route could be a valuable technique not only in an AED but also for breakthrough pain by offering a fast onset of pain control in moderate to severe painful conditions. It could also provide pain relief and allow topical anaesthetics to take effect on the skin prior to IV establishment. Therefore this may be a suitable alternative to IND.

A similar and more recent double blinded RCT trial by Saunders et al (2007) claimed efficacy of a larger dosing regimen with a mean dose of 2µg/kg INF (50µg/ml) for pain reductions in paediatric orthopaedic trauma compared with IVM at 0.1mg/kg in 60 3-12 year old children. This study reports positive outcome for INF following both patients and carers reporting very effective pain management and satisfaction using this treatment method. However there is little information in the paper of methodology and results are given in percentages’ rather than a P value or NTT which should be expected in a rigorous creditable RCT of two agents (Bowling & Ebrahim 2005) reducing the level of evidence applied to the paper to L3.

Even supposing the results are an accurate reflection of the efficacy and safety of INF, particularly the fact that no significant difference in pain score or side effect profile and INF is a way forward, the lack of detail the randomisation process and analysis of data in the study methodology merely implies that these results maybe flawed. Interestingly given the concentration of fentanyl 50µg/ ml a dosing volume for a 25kg child would have required one ml = 0.5ml per-nostril therefore suggesting some of the administration may have been oral rather than IN and present the issues of bad taste which is put forward as a possible study limitation by the authors. Then again there are no complications or reports on taste presented in the results and the author’s conclusion on the efficacy of INF for acute pain management in children may be founded. However, without sourcing more details from the authors it cannot be considered evidence to inform this dissertations aims but merely an example of poor research or appropriate omission by publishers. Further suggesting there remains a requirement for more research on the topic within double blind, equivalence, RCT focused on INF efficacy and dosing with sound methodology that is transparent in publication to answer the dissertation question.

Conversely an older and more rigorous study which also looked at dose related analgesic effect between routes of administration is by Manjushree et al (2002). The authors demonstrated the clinical efficacy of INF in a cohort of 32 children (aged 4-8 yrs) in a postoperative situation and with a double blind level 1 RCT. The study design gives the impression of sound methodology as blinding, assessment and analysis of data was appropriate and available for scrutiny in the paper, particularly the analysis of both nonparametric and nominal data. The only weakness is possibly the sample size of 32 patients. Although the authors performed a power calculation which identified 40 patients to show a significant affect, they only recruited 32 patients, furthermore, this appears to be an equivalency study where the authors hypothesised INF would be equal to and not inferior to IVF therefore would have needed a larger sample to de