RESEARCH REPORT
`
`doi:10.1111/j.1360-0443.2009.02724.x
`
`Randomized controlled trial comparing the
`effectiveness and safety of intranasal and
`intramuscular naloxone for the treatment of
`suspected heroin overdoseadd_2724 2067..2074
`
`Debra Kerr1,2, Anne-Maree Kelly2,3, Paul Dietze4,5, Damien Jolley5 & Bill Barger6
`Victoria University, School of Nursing and Midwifery, St Albans,Victoria, Australia,1 University of Melbourne, Parkville,Victoria, Australia,2 Joseph Epstein Centre for
`Emergency Medicine Research, Sunshine Hospital, St Albans,Victoria,Australia,3The Macfarlane Burnet Institute for Medical Research and Public Health, Melbourne,
`Victoria, Australia,4 Monash Institute of Health Services Research, Clayton,Victoria, Australia5 and Ambulance Victoria, Doncaster,Victoria, Australia6
`
`ABSTRACT
`
`Aims Traditionally, the opiate antagonist naloxone has been administered parenterally; however, intranasal (i.n.)
`administration has the potential to reduce the risk of needlestick injury. This is important when working with popu-
`lations known to have a high prevalence of blood-borne viruses. Preliminary research suggests that i.n. administration
`might be effective, but suboptimal naloxone solutions were used. This study compared the effectiveness of concentrated
`(2 mg/ml) i.n. naloxone to intramuscular (i.m.) naloxone for suspected opiate overdose. Methods This randomized
`controlled trial included patients treated for suspected opiate overdose in the pre-hospital setting. Patients received 2 mg
`of either i.n. or i.m. naloxone. The primary outcome was the proportion of patients who responded within 10 minutes
`of naloxone treatment. Secondary outcomes included time to adequate response and requirement for supplementary
`naloxone. Data were analysed using multivariate statistical techniques. Results A total of 172 patients were enrolled
`into the study. Median age was 29 years and 74% were male. Rates of response within 10 minutes were similar: i.n.
`naloxone (60/83, 72.3%) compared with i.m. naloxone (69/89, 77.5%) [difference: -5.2%, 95% confidence interval
`(CI) -18.2 to 7.7]. No difference was observed in mean response time (i.n.: 8.0, i.m.: 7.9 minutes; difference 0.1, 95%
`CI -1.3 to 1.5). Supplementary naloxone was administered to fewer patients who received i.m. naloxone (i.n.: 18.1%;
`i.m.: 4.5%) (difference: 13.6%, 95% CI 4.2–22.9). Conclusions Concentrated intranasal naloxone reversed heroin
`overdose successfully in 82% of patients. Time to adequate response was the same for both routes, suggesting that the
`i.n. route of administration is of similar effectiveness to the i.m. route as a first-line treatment for heroin overdose.
`
`Keywords Heroin, intranasal, naloxone, opioid, overdose, resuscitation.
`
`Correspondence to: Debra Kerr, Victoria University, School of Nursing and Midwifery, Building 4, McKechnie Street, St Albans, Vic. 3021, Australia.
`E-mail: deb.kerr@vu.edu.au
`Submitted 4 May 2009; initial review completed 17 June 2009; final version accepted 1 July 2009
`
`INTRODUCTION
`
`Heroin overdose is a major cause of death in some coun-
`tries [1–4]. In most instances, timely treatment with
`naloxone, an opiate antagonist, reverses opioid toxicity.
`In the community setting, paramedics administer nalox-
`one routinely for suspected opioid overdose via the intra-
`muscular (i.m.) and/or intravenous (i.v.) routes [5–7].
`Administration of the drug by these routes to populations
`such as injecting drug users carries some risk. Injecting
`drug users are often infected with blood-borne viruses
`
`such as human immunodeficiency virus (HIV), hepatitis
`B (HBV) and hepatitis C (HCV) [8–10], and in spite of
`best practice guidelines designed to minimize needlestick
`injury among health workers, needlestick injuries occur,
`allowing for the possibility of blood-borne virus trans-
`mission. Among health care workers, 4% of HIV infec-
`tions and 40% of HBV and HCV infections occur after
`occupational exposure [11].
`There is growing interest in intranasal (i.n.) ad-
`ministration of naloxone [12–17]. The benefits of
`i.n.
`administration include ease of access, greatly reduced
`
`© 2009 The Authors. Journal compilation © 2009 Society for the Study of Addiction
`
`Addiction, 104, 2067–2074
`
`Nalox1036
`Nalox-1 Pharmaceuticals, LLC
`Page 1 of 8
`
`

`

`2068
`
`Debra Kerr et al.
`
`needlestick injury risk and the potential for peer and non-
`health professional administration. Its use in acute over-
`dose is supported by a number of small cohort studies
`[18–22]. To date, there has only been one randomized
`trial comparing i.n. and i.m. administration [22]. It found
`i.m. administration resulted in shorter response time
`than i.n. administration (mean 6 minutes versus 8
`minutes), but the i.n. route was successful for 74% of
`patients. The preparation used for i.n. administration in
`that study (2 mg in 5 ml) far exceeded recommendations
`for i.n. use of drugs that specify volumes of less than 1 ml
`per nostril [12]. It was, however, the only preparation
`available at the time of
`that study. That raised the
`question of whether concentrated, small-volume dosing
`would improve the effectiveness of i.n. naloxone.
`The aim of this study was to determine the effective-
`ness and safety of concentrated (2 mg/ml) i.n. naloxone
`compared to i.m. naloxone for treatment of suspected
`opiate overdose in the pre-hospital setting. Specifically,
`the study sought to compare the two preparations in
`terms of response times, side effects, need for a second
`dose of naloxone and final outcomes.
`
`METHODS
`
`Participants
`
`This was a prospective, randomized, unblinded trial con-
`ducted in Melbourne, Victoria, Australia. Patients requir-
`ing treatment by six designated branches of Metropolitan
`Ambulance Service (MAS, Victoria) for suspected opiate
`overdose during the period from 1 August 2006 to 31
`January 2008 were considered for enrolment. We chose
`these branches as they were located in areas with higher
`incidence of heroin overdose, known historically to
`capture more than half of the heroin overdoses in the
`metropolitan region [23].
`Patients were eligible for enrolment if they suffered a
`suspected opiate overdose [altered conscious state, pin-
`point pupils, respiratory depression (respirations < 10)],
`were unrousable as defined by Glasgow Coma Score (GCS)
`ⱕ12 and had no major facial trauma, blocked nasal
`passages or epistaxis. The GCS score was chosen as the
`measure of sedation because it is the parameter used
`operationally in the ambulance service within which our
`study was conducted [24].
`We were aiming for a consecutive sample. However,
`paramedic staff turnover meant that not all eligible
`patients were enrolled during the study period. Paramed-
`ics required training in the study protocol and use of
`the atomization device before enrolling participants. This
`meant that potential participants, who were treated by
`paramedics who had not been trained, could not be
`enrolled into the study. During the study period there
`
`were approximately 1300 heroin overdose attendances,
`defined as a patient with a positive response to the admin-
`istration of naloxone by paramedics, in metropolitan
`Melbourne [25].
`Melbourne Health Human Research Ethics Com-
`mittee (HREC) approved the study. Requirement for
`individual patient consent was waived. Subjects were
`informed of their participation by way of an information
`letter after regaining consciousness which allowed them
`to withdraw themselves from the study or seek further
`information.
`
`Procedure
`
`Allocation of mode of administration (i.n. or i.m.) was
`achieved by block randomization using an online com-
`puter program to achieve a random sequence of alloca-
`tions. Block randomization was performed to achieve
`equal distribution of allocations (i.n. or i.m.) to each
`study site. The nature of pre-hospital emergency care
`and the urgency of treatment for this condition prohi-
`bits more sophisticated double-treatment randomization
`techniques.
`Randomization envelopes, present in each ambu-
`lance, were designed by the study investigators to conceal
`the randomization group. The allocation notice was
`positioned between the study information sheet and the
`envelope was made of thicker, non-transparent paper.
`This was designed to prevent paramedics choosing the
`randomization arm selectively for potential subjects. All
`envelopes were identical from the outside. All envelopes
`were numbered sequentially according to the block ran-
`domization procedure, and all envelopes were accounted
`for at monthly intervals and at the end of the study.
`After determining eligibility, a randomization enve-
`lope was opened at the scene, allocating patients to
`receive either i.n. naloxone 2 mg or i.m. naloxone 2 mg.
`Supportive care (primarily breathing support) was
`administered simultaneously, in accordance with ambu-
`lance clinical practice guidelines for this condition.
`Administration by i.m. injection was by standard MAS
`practice using a pre-packaged ‘min-i-jet’™ preparation
`containing naloxone solution (2 mg/5 ml). Naloxone for
`i.n. administration was constituted in a tamper-evident
`vial as a preparation of 2 mg in 1 ml, manufactured
`specifically for the study and complying with national
`medication quality and safety standards. At the scene,
`contents of the vial were withdrawn into a luer-lock
`syringe, and the syringe was then attached to a mucosal
`atomization device (MAD®). Paramedics were instructed
`to depress the syringe rapidly during i.n. administration
`to achieve adequate atomisation. Study participants
`received 1 mg (0.5 ml) in each nostril.
`Standard supportive care,
`including airway and
`breathing support as needed, continued throughout the
`
`© 2009 The Authors. Journal compilation © 2009 Society for the Study of Addiction
`
`Addiction, 104, 2067–2074
`
`Nalox1036
`Nalox-1 Pharmaceuticals, LLC
`Page 2 of 8
`
`

`

`data collection period until either recovery or transport to
`hospital. All patients who failed to respond to either form
`of naloxone treatment after 10 minutes were eligible for
`a ‘rescue’ dose of 0.8 mg i.m. naloxone. The 10-minute
`recommendation was chosen for consistency with treat-
`ment recommendations already laid down in the relevant
`ambulance service protocols [26].
`
`Measurements
`
`Paramedics entered study information into an electronic
`patient case record (e-PCR), as per the Victorian Ambu-
`lance Clinical Information System (VACIS). The e-PCR
`is the tool used by paramedics to document emergency
`care administered for all cases. The data for this study
`were extracted by explicit review of these files. Informa-
`tion collected included demographic data [age, gender,
`vital signs (including respiratory rate, pulse, GCS)], sus-
`picion of other drugs/alcohol taken, specific location,
`other people present, resuscitative measures (basic life
`support, airway management), naloxone administration
`(dose, route, time of administration, difficulty during
`administration, requirement for secondary naloxone),
`response times, side effects and final outcome (self-care,
`hospitalization, death). Data were entered directly into a
`Microsoft Access database developed specifically for this
`study. All data entries were checked for accuracy by
`an independent blinded research assistant. A third
`researcher arbitrated discrepant data extraction (three
`cases only).
`The primary outcome of interest was the proportion
`of patients with an adequate response within 10 minutes
`of naloxone administration. Response was defined as
`effective and spontaneous respirations at a rate ⱖ 10
`per minute and/or GCS ⱖ 13. Patients who received a
`supplementary dose were classified automatically as
`not achieving an adequate response within 10 minutes.
`This end-point was chosen to be consistent with current
`ambulance practice guidelines, where secondary nalox-
`one is recommended for inadequate response after a
`10-minute period [25]. While, for many clinicians, rever-
`sal of respiratory depression is the key outcome, improve-
`ment in level of consciousness, indicating the reversal of
`over-sedation responsible for respiratory depression, has
`been used by previous studies in this field [18,19] as an
`indicator of successful treatment.
`Secondary outcomes included time to adequate
`response, hospitalization, adverse event rate and require-
`ment for ‘rescue’ naloxone due to inadequate primary
`response as judged by the treating paramedics.
`Adverse events were grouped into three categories
`including drug-related (vomiting, nausea, seizure, sweat-
`ing, tremor, acute pulmonary oedema, increased blood
`pressure, tremulousness, seizures, ventricular tachycar-
`
`Intranasal naloxone for suspected heroin overdose
`
`2069
`
`dia and fibrillation, cardiac arrest, agitation and paraes-
`thesia), administration-related (nasal obstruction, nasal
`deformity) and study-related (epistaxis, ruptured septum,
`spitting, coughing,
`leakage of
`solution from nasal
`passages).
`
`Data analyses
`
`Descriptive analyses [proportion, mean, median, effect
`size difference with 95% confidence interval (CI)] were
`conducted using Intercooled Stata version 8.2 [27] to
`describe the demographic data and compare groups (i.n.
`and i.m.) for observed differences (drug use, alcohol use).
`Primary outcomes were compared by univariate analysis
`including observed difference and odds ratio (OR) with
`95% CI, hazard ratio (HR) and c2 analysis. Correlates
`included in the multivariate models (logistic regression,
`Cox regression) were age, gender and concomitant
`alcohol and/or drug use.
`Response time was compared using Kaplan–Meier
`survival analysis. A clinically significant difference in
`response time was defined as 1 minute. This end-point
`was based on the likelihood of oxygen de-saturation after
`1 minute as a result of respiratory depression. For all
`patients, entry time was defined as 1 minute after admin-
`istration by either route; exit time was the earliest of (i)
`adequate response; or (ii) rescue naloxone; or (iii) last
`recorded observation. Only the first of these exit times
`was regarded as an event, and the latter two were con-
`sidered as censored observations.
`Based on previous studies [18,19,22], we needed to
`recruit at least 84 patients per group to detect a difference
`in proportions for successful response to naloxone treat-
`ment of 11% (100% versus 89%) with power 80% (Inter-
`cooled Stata version 10.0) [28]. With this sample, and
`assuming similar results of around 95% success for both
`groups, the width of the 95% CI for difference in risk will
`be ⫾ 6.4%.
`
`RESULTS
`
`Two hundred and sixty-six patients were treated for
`suspected heroin overdose at the enrolment sites during
`the study period; 13 patients were not considered for
`study enrolment. A further 75 patients were not eligible,
`as shown in the participant flow diagram (Fig. 1), includ-
`ing 20 patients who could not be included because
`paramedics at the site had not been trained in the study
`protocol. Of the remaining 178 patients, six patients were
`excluded from participation for the following reasons:
`equipment for intranasal administration was missing for
`three patients and three patients became alert prior to
`naloxone administration (two in the i.n. group and one in
`the i.m. group). These six patients were excluded from
`
`© 2009 The Authors. Journal compilation © 2009 Society for the Study of Addiction
`
`Addiction, 104, 2067–2074
`
`Nalox1036
`Nalox-1 Pharmaceuticals, LLC
`Page 3 of 8
`
`

`

`2070
`
`Debra Kerr et al.
`
`Total patients administered naloxone
`during study period at study sites
`(n=266)
`
`Not considered for study enrolment
`(n=13)
`
`Total patients assessed for eligibility
`(n=253)
`
`Excluded (n=75)
`-Not meeting study criteria (n=55)
`-Paramedic not trained in protocol (n=20)
`
`Randomised
`(n=178)
`
`Allocated to i.n. (n=88)
`-Became alert (n=2)
`-Equipment missing (n=3)
`
`Allocated to i.m. (n=90)
`-Became alert (n=1)
`
`Received i.n. (n=83)
`
`Received i.m. (n=89)
`
`Figure 1 Participant flow diagram. i.m.: intramuscular; i.n.: intranasal
`
`final data analysis. Hence, data were not analysed on
`an ‘intention-to-treat’ basis but, rather, analysed by the
`treatment they received.
`The final sample consisted of 172 patients who
`received i.n. (83 patients) or i.m. (89 patients) naloxone.
`The characteristics of
`the patients are shown in
`Table 1 according to their allocated treatment. Patients
`were broadly similar for age, gender and treatment time.
`The median age was 29 years, and 74% were male.
`An important difference in baseline characteristics was
`observed, with more patients in the i.n. group suspected
`of concomitant drug use compared to the i.m. group
`[i.n.: 21.7%, i.m.: 9.0%, difference 12.7% (95% CI 2.0,
`23.4)].
`Study outcomes are shown in Table 2. One hundred
`and twenty-nine patients (75%) achieved an adequate
`response within 10 minutes from initial naloxone treat-
`ment, 60 (72.3%) in the i.n. group and 69 (77.5%) in the
`i.m. group [difference -5.2% (95% CI -18.2, 7.7%)].
`Mean response time (minutes) was similar between the
`two groups [i.n.: 8.0, i.m.: 7.9, HR 0.8 (95% CI 0.6, 1.2)],
`as shown in Fig. 2. The absence of significant difference
`
`Table 1 Comparison of characteristics for patients treated for
`heroin overdose with intranasal or intramuscular naloxone.
`
`Variable
`
`Age (mean years)
`Treatment timea (mean minutes)
`Male
`Concomitant alcohol
`Concomitant drugs
`Concomitant alcohol ⫾ drugs
`Public use
`
`Intranasal
`(%)
`n = 83
`
`30.6
`13.1
`64 (77.1)
`25 (30.1)
`18 (21.7)
`39 (47.0)
`42 (50.6)
`
`Intramuscular
`(%)
`n = 89
`
`31.8
`13.4
`63 (70.8)
`31 (34.8)
`8 (9.0)b
`33 (37.1)
`47 (52.8)
`
`aTime from ambulance call to administration of naloxone treatment.
`bObserved difference 12.7% (95% confidence interval 2.0, 23.4).
`
`was supported by multivariate analysis for adequate
`response within 10 minutes [OR 0.7 (95% CI 0.3, 1.5)]
`and actual response time [HR 0.84 (95% CI 0.6, 1.2)].
`Rescue naloxone was administered more often to
`patients in the i.n. group (18.1%) compared with those
`
`© 2009 The Authors. Journal compilation © 2009 Society for the Study of Addiction
`
`Addiction, 104, 2067–2074
`
`Nalox1036
`Nalox-1 Pharmaceuticals, LLC
`Page 4 of 8
`
`

`

`Intranasal naloxone for suspected heroin overdose
`
`2071
`
`Table 2 Comparison of outcomes for patients treated by intranasal (i.n.) or intramuscular (i.m.) naloxone.
`
`Outcome
`
`i.n. (83)
`n (%)
`
`i.m. (89)
`n (%)
`
`Difference (95% CI)
`
`Univariate
`analysis
`OR (95% CI)
`
`Multivariate
`analysis
`OR (95% CI)
`
`60 (72.3) 69 (77.5) -5.2%, (-18.2, 7.7)
`Adequate response ⱕ 10 minutes
`0.7, (0.3, 1.5)
`0.8, (0.4, 1.5)
`4.8, (1.4, 16.3)*
`4.7, (1.6, 14.1)
`Rescue naloxone for inadequate response 15 (18.1)
`4 (4.5)
`13.6%, (4.2, 22.9)
`3.1%, (-10.3, 16.4) 1.2, (0.6, 2.3)
`1.3, (0.6, 2.7)
`Hospitalization
`24 (28.9) 23 (25.8)
`0.2%, (-11.6, 11.9) 1.0, (0.5, 2.2)
`1.1, (0.5, 2.5)
`Minor adverse event
`16 (19.3) 17 (19.1)
`HR (95% CI)
`HR (95% CI)
`0.8, (0.6, 1.2)** 0.84, (0.6, 1.2)***
`
`Mean response time (minutes)
`
`8.0
`
`7.9
`
`0.1 (-1.3, 1.5)
`
`*P = 0.01; **P = 0.29; ***P = 0.29. HR: hazard ratio in i.n. group, relative to i.m. group; OR: odds ratio for each outcome in i.n. group, relative to i.m.
`group; CI: confidence interval.
`
`1.00
`
`0.75
`
`0.50
`
`0.25
`
`0.00
`
`Proportion
`without
`response
`
`0
`
`10
`
`20
`
`30
`
`40
`
`Response Time
`
`i.m.
`
`i.n.
`
`Figure 2 Kaplan–Meier survival curve comparing response times for patients who receive intranasal (i.n.) or intramuscular (i.m.) naloxone
`
`in the i.m. group (4.5%) [difference 13.6% (95% CI 4.2,
`22.9%)]. After controlling for age, gender and suspected
`concomitant alcohol and/or drugs,
`this difference
`remained statistically significant [OR 4.8 (95% CI 1.4,
`16.3)]. Twenty-four patients did not achieve an adequate
`response at 10 minutes and were not administered
`secondary naloxone (i.n.: 8/23, i.m.: 16/20). Average
`response from initial naloxone treatment was 16 minutes
`for these cases. It is our assumption that paramedics
`chose to wait for a response after the 10-minute cut-off,
`and patients responded without secondary naloxone
`administration. However, we did not collect information
`regarding reasons for not administering naloxone for
`these cases.
`
`There was one major adverse event. A patient who
`received i.m. naloxone had a grand mal epileptic seizure,
`was given i.v. diazepam, and was transferred subse-
`quently to hospital
`for further management. Minor
`adverse events were similar between the two groups
`(i.n.: 19.3%, i.m.: 19.1%; difference 0.2% 95% CI -11.6,
`11.9), as were hospitalization rates (i.n.: 28.9%, i.m.:
`25.8%; difference 3.1% 95% CI -10.3, 16.4). No differ-
`ence was observed in agitation and/or violence (i.n.:
`6.0%, i.m.: 7.9%), nausea and/or vomiting (i.n.: 8.4%,
`i.m.: 7.9%) and headache (i.n.: 4.8%, i.m.: 3.3%) after
`naloxone treatment. To our knowledge there were no
`needlestick injuries during i.m. administration of nalox-
`one during the study period.
`
`© 2009 The Authors. Journal compilation © 2009 Society for the Study of Addiction
`
`Addiction, 104, 2067–2074
`
`Nalox1036
`Nalox-1 Pharmaceuticals, LLC
`Page 5 of 8
`
`

`

`2072
`
`Debra Kerr et al.
`
`DISCUSSION
`
`Emergency medical service (EMS) personnel are at an
`increased risk of blood-borne virus exposure when pro-
`viding treatment to injecting drug users, a population
`with an increased prevalence of HIV, and HBV and
`HBC [29–31]. Administration of medication via non-
`parenteral routes is one means of reducing needlestick
`injury risk. This study has shown that administration
`of naloxone via the i.n. route, using a concentrated
`solution, to patients with suspected heroin overdose in
`the pre-hospital setting is a safe and effective treatment
`option, with similar response rates, response times and
`side-effect profile to i.m. administration.
`Previous studies have reported success rates for i.n.
`naloxone between 74 and 91% [18,19,22]. In these
`studies, successful treatment was defined as an adequate
`response to i.n. naloxone without
`the requirement
`to administer secondary naloxone treatment. Taken
`together with this study, they provide strong evidence
`that i.n. naloxone is effective for initial treatment of
`heroin overdose in the community.
`Current ambulance protocols for naloxone in most
`jurisdictions recommend i.m. administration [26,32].
`The protocol for the ambulance service involved in this
`study involves naloxone administration using a pre-
`packaged syringe and needle (min-i-jetTM), which means
`that needlestick injury protection is reliant upon para-
`medics adhering to good practice around the manage-
`ment of needles; i.n. administration of naloxone offers
`clear advantages here in terms of a reduction in needle-
`stick injury risk. Given our findings, it would appear
`that i.n. naloxone is a viable therapy that reduces the
`possibility of needlestick injury among paramedics when
`compared to parenteral alternatives.
`While the finding that approximately a quarter of
`patients in each group did not respond to naloxone
`is important, it should be noted that there was no statis-
`tically significant difference between the groups with
`regard to the proportion of non-responders. Lack of
`response to naloxone therapy after ambulance response
`has been reported (20–63%) [18,19,22]. Non-response
`may reflect simple misclassification (heroin overdose is
`notoriously difficult to define) [33], but may reflect other
`causes such as the possibility that the delay between
`overdose and the attendance of the ambulance reduces
`adequate response, with greater delays possibly being
`associated with more advanced respiratory depression.
`Polydrug use and other physical comorbidity may also
`be relevant [34]. Irrespectively, the non-response we
`observed highlights the importance of pre-hospital sup-
`portive care (by bystanders followed initially by EMS
`personnel) that remains an essential component in
`preventing deaths.
`
`Response to i.m. naloxone treatment was slower in
`this study (8 minutes) in comparison to previous research
`(6 minutes) [22]. It is unclear why this is so, as the nalox-
`one preparation and protocol
`for i.m. administration
`were identical in both studies, but there may have been
`differences between studies regarding the type and quan-
`tity of drugs used by participants prior to overdose.
`Response to i.n. administration was the same as reported
`previously [22], despite the change in concentration.
`A concentrated preparation of naloxone has not
`been investigated previously. For optimal absorption and
`effectiveness, it is advised that medication for i.n. admin-
`istration be prepared in volumes of less than 1 ml per
`nostril [12]. A suitable preparation for nasal administra-
`tion (<1 ml per nostril) of a dose equivalent to that used
`in this study is not currently available in Australia or
`overseas. Naloxone for i.n. administration was manu-
`factured specifically for this study under the legislative
`authority as a registered clinical trial. Previous studies
`using dilute preparations have reported success rates
`between 74 and 91% [18,19,22]. The success rate in
`this study is not significantly better than these, so it
`cannot be concluded that the concentrated solution is
`more effective. That said, smaller volumes are easier to
`administer and lend themselves more effectively to pre-
`packaged devices. In addition, there were no reports
`of excess fluid expulsion from the nose or coughing by
`study subjects in this current study, as was observed
`in previous research [22].
`Although patients who received i.n. naloxone were
`4.8 times (95% CI 1.4, 16.3) more likely to receive rescue
`naloxone, this finding needs to be considered from a cli-
`nical perspective. Administration of rescue naloxone to
`patients included in our study was a subjective decision
`made by paramedics at the scene, and was very depen-
`dent upon the individual paramedic and their comfort
`waiting for an adequate response, the patient’s respira-
`tory and conscious state and patient request for further
`naloxone. Paramedics were encouraged to administer
`secondary naloxone if an inadequate response was
`observed after 10 minutes. It is possible that a response
`might have been observed for some patients if a longer
`observation period had occurred. Also, randomization
`was not blinded. A double-blind study design would
`have eliminated this limitation. Paramedics might have
`administered secondary naloxone to patients who
`received the i.n. allocation due to apprehension about the
`effectiveness of the i.n. treatment option. However, the
`possibility that patients who receive i.n. naloxone may
`require rescue naloxone more often cannot be ruled
`out by our study.
`The fact that 72% of the i.n. group responded within
`10 minutes highlights the potential of i.n. naloxone to
`be used for peer administration. Naloxone distribution
`
`© 2009 The Authors. Journal compilation © 2009 Society for the Study of Addiction
`
`Addiction, 104, 2067–2074
`
`Nalox1036
`Nalox-1 Pharmaceuticals, LLC
`Page 6 of 8
`
`

`

`programmes using parenteral naloxone have been insti-
`tuted in some places [32,35], and favourable reports of
`lives saved have been reported [35]. The preferred route
`for peer naloxone administration is an important issue,
`and has been reported in a separate study [36]. Nasal
`administration for peer naloxone distribution was pre-
`ferred (74%) by current heroin users (n = 99) in a study
`performed in Melbourne (Australia) during 2007 [36].
`Administration via the i.n. route may be a simpler option
`for those without professional health care training and
`largely eliminates infection risk. An opioid overdose pre-
`vention programme in Boston (USA) distributes an intra-
`nasal naloxone spray to potential bystanders [37]. They
`report that after 15 months from programme commence-
`ment there have been 74 successful overdose reversals,
`and few problems with the i.n. spray.
`Our study responds to the need for well-designed
`randomized clinical trials in the drug and emergency
`medicine research fields. It does, however, have some
`limitations that should be considered when interpreting
`the results. The study may have been strengthened by a
`double-blinded study design; however, the pre-hospital
`setting for
`research poses challenges
`that
`require
`flexibility and simplicity in study design [38]. Not all
`patients were enrolled into the study, although we
`encouraged paramedics to consider all patients treated
`for heroin overdose for recruitment. Our study did not
`include all ambulance sites in metropolitan Melbourne,
`hence only 266 were considered for recruitment. This
`might have resulted in a systematic bias in enrolment.
`We were also unable to measure for opioid, polydrug or
`alcohol load. Hence, heroin overdose was not confirmed.
`Tolerance to heroin has been shown to be influenced
`greatly by alcohol and polydrug use [39–41]. Paramed-
`ics document routinely evidence of polydrug and/or
`alcohol consumption prior to the event, but there may
`have been some unidentified cases. Our sample size cal-
`culations were made on data that was available at the
`time of study design. This over-estimated significantly
`the success rates of both routes of administration and
`posed a potential threat to the study’s power. This
`is countered by the almost identical response times,
`so a clinically significant difference in effectiveness is
`unlikely.
`In conclusion, we have shown that naloxone admin-
`istered via the i.n. route is an effective and safe interven-
`tion for the initial management of heroin overdose.
`However, the concentrated preparation we used was not
`more effective than the less concentrated version used in
`a previous study. The i.n. option offers rescuers a needle-
`less option as first-line treatment and opens opportunities
`for wider distribution of naloxone for peer and non-
`health care administration. A low adverse event rate was
`found for both (i.n. and i.m.) routes.
`
`Intranasal naloxone for suspected heroin overdose
`
`2073
`
`Declarations of interest
`
`None.
`
`Acknowledgements
`
`We would like to acknowledge Kerry Leigh, a paramedic
`of the Metropolitan Ambulance Service, for her consider-
`able efforts in training paramedics of all recruiting sites,
`and coordination of equipment required for the study
`and People Strategy Innovation Pty Ltd for research
`support services. This study was supported by a grant
`received from the Drug Policy and Services, Department
`of Human Services, Melbourne, Victoria, Australia. No
`restrictions were imposed on the investigators. The
`design and conduct of the study; collection, manage-
`ment, analysis and interpretation of the data; and prepa-
`ration, review and approval of the manuscript was the
`responsibility of the authors. The funders held no respon-
`sibility for these tasks. Associate Professor Paul Dietze
`is funded by a Career Development Award from the
`National Health Medical Research Council (NHMRC)
`(Australia) Grant. The study was registered with the
`‘Australian New Zealand Clinical Trials Registry’
`(ACTRN: 12606000322538).
`
`References
`
`1. Bryant W. K., Galea S., Tracy M., Markham Piper T., Tardiff
`K. J., Vlahov D. Overdose deaths attributed to methadone
`and heroin in New York City, 1990–1998. Addiction 2004;
`99: 846–54.
`2. Hall W. D., Degenhardt L. J., Lynskey M. T. Opioid overdose
`mortality in Australia, 1964–1997: birth-cohort trends.
`Med J Aust 1999; 171: 34–7.
`3. Hickman M., Madden P., Henry J., Baker A., Wallace C.,
`Wakefield J. et al. Trends in drug overdose deaths in England
`and Wales 1993–98: methadone does not kill more people
`than heroin. Addiction 2003; 98: 419–25.
`4. Preti A., Miotto P., De Coppi M. Deaths by unintentional
`illicit drug overdose in Italy, 1984–2000. Drug Alcohol
`Depend 2002; 66: 275–82.
`5. Darke S., Williamson A., Ross J., Teesson M. Non-fatal
`heroin overdose, treatment exposure and client characteris-
`tics: findings from the Australian treatment outcome study
`(ATOS). Drug Alcohol Rev 2005; 24: 425–32.
`6. Buajordet I., Naess A. C., Jacobsen D., Brors O. Adverse
`events after naloxone treatment of episodes of suspected
`acute opioid overdose. Eur J Emerg Med 2004; 11: 19–23.
`7. Sporer K. A., Firestone J., Isaacs S. M. Out-of-hospital treat-
`ment of opioid overdoses in an urban setting. Acad Emerg
`Med 1996; 3: 660–7.
`8. Crofts N., Jolley D., Kaldor J., van Beek I., Wodak A. Epide-
`miology of hepatitis C virus infection among injecting drug
`users in Australia. J Epidemiol Commun Health 1997; 51:
`692–7.
`9. Davoli M., Perucci C. A., Rapiti E., Bargagli A. M., D’Ippoliti
`D., Forastiere F. et al. A persistent rise in mortality among
`injection drug users in Rome, 1980 through 1992. Am J
`Public Health 1997; 87: 851–3.
`
`© 2009 The Authors. Journal compilation © 2009 Society for the Study of Addiction
`
`Addiction, 104, 2067–2074
`
`Nalox1036
`Nalox-1 Pharmaceuticals, LLC
`Page 7 of 8
`
`

`

`2074
`
`Debra Kerr et al.
`
`10. Kaplan E. H., Heimer R. A model-based estimate of HIV
`infectivity via needle sharing. J Acquir Immune Defic Syndr
`1992; 5: 1116–8.
`11. World Health Organization (WHO). World Health Report.
`Geneva: WHO; 2002.
`12. Wolfe T. R., Bernstone T. Intranasal drug