Drug and Alcohol Review (July 2009), 28, 347–352
`DOI: 10.1111/j.1465-3362.2009.00057.x
`
`Overdose deaths following previous non-fatal heroin overdose:
`Record linkage of ambulance attendance and death registry data
`
`MARK A. STOOVÉ1, PAUL M. DIETZE1 & DAMIEN JOLLEY2
`
`1Centre for Epidemiology and Population Health Research, Macfarlane Burnet Institute for Medical Research and Public
`Health, Melbourne, Australia, and 2Monash Institute of Health Services Research, Faculty of Medicine, Nursing and
`Health Sciences, Monash University, Melbourne, Australia
`
`Abstract
`Introduction and Aims. Experiencing previous non-fatal overdoses have been identified as a predictor of subsequent
`non-fatal overdoses; however, few studies have investigated the association between previous non-fatal overdose experiences and
`overdose mortality. We examined overdose mortality among injecting drug users who had previously been attended by an
`ambulance for a non-fatal heroin overdose. Design and Methods. Using a retrospective cohort design, we linked data on
`non-fatal heroin overdose cases obtained from ambulance attendance records in Melbourne, Australia over a 5-year period
`(2000–2005) with a national death register. Results. 4884 people who were attended by ambulance for a non-fatal heroin
`overdose were identified. One hundred and sixty-four overdose deaths occurred among this cohort, with an average overdose
`mortality rate of 1.20 per 100 person-years (95% CI, 1.03–1.40). Mortality rate decreased 10-fold after 2000 coinciding with
`widely reported declines in heroin availability. Being male, of older age (>35 years) and having been attended multiple times
`for previous non-fatal overdoses were associated with increased mortality risk. Discussion and Conclusions. As the first
`to show a direct association between non-fatal overdose and subsequent overdose mortality, this study has important implications
`for the prevention of overdose mortality. This study also shows the profound effect of macro-level heroin market dynamics on
`overdose mortality. [Stoové MA, Dietze PM, Jolley D. Overdose deaths following previous non-fatal heroin overdose:
`Record linkage of ambulance attendance and death registry data. Drug Alcohol Rev 2009;28:347–352]
`
`Key words: overdose, mortality, injecting drug use, cohort study, data linkage.
`
`Introduction
`
`The health consequences of injecting drugs over an
`extended time period are profound. Dependent inject-
`ing drug users (IDU) are of considerable risk of mor-
`bidity [1], and substantially greater risk of mortality
`(estimated at between six and 20 times) compared with
`their non-injecting peers [2,3]. Over the past two
`decades heroin overdose has emerged as a major public
`health challenge. Opioid use constitutes the largest con-
`tributor to illicit drug deaths, driven largely by the
`injection of heroin [4,5].
`Mortality among IDU is dependent on the complex
`interaction of precursors and sequelae associated with
`injecting drugs. Factors identified as increasing mortal-
`ity risk among IDU include: drug toxicity, dose and
`
`frequency of injection [6,7]; psychosocial and environ-
`mental factors, such as imprisonment and homeless-
`ness
`[8,9]; medical complications
`resulting from
`injecting drugs [10,11]; and opioid pharmacotherapies
`(primarily during
`initial
`treatment periods;
`see
`[12,13]). In addition, demographic characteristics,
`such as being male and older age or longer duration of
`injecting, have been shown to be associated with greater
`overdose risk [1,2].
`Fatal overdose, however, makes up only a small pro-
`portion of overdose events, estimated at between 2%
`and 4% [14]. Others have reported annual rates of
`non-fatal overdose among IDU of between 10% and
`30% [15–19]. Non-fatal overdose results in significant
`morbidity [19,20] and recent studies have identified
`previous non-fatal overdose experiences as a significant
`
`Mark A. Stoové PhD, Paul M. Dietze PhD, Damien Jolley MAE. Correspondence to Dr Mark A. Stoové, Centre for Epidemiology and Population
`Health Research, Macfarlane Burnet Institute for Medical Research and Public Health, PO Box 2284, Melbourne, Vic. 3001, Australia. Tel: +61
`(0) 3 8506 2301; Fax: +61 (0) 3 9282 2138; E-mail: stoove@burnet.edu.au
`
`Received 16 May 2008; accepted for publication 12 August 2008.
`
`© 2009 Australasian Professional Society on Alcohol and other Drugs
`
`Opiant Exhibit 2023
`Nalox-1 Pharmaceuticals, LLC v. Opiant Pharmaceuticals, Inc.
`IPR2019-00690
`Page 1
`
`

`

`348
`
`M. A. Stoové et al.
`
`predictor of subsequent non-fatal overdose [17,21,22].
`However, to our knowledge, researchers are yet to
`report associations between previous non-fatal over-
`dose and subsequent overdose mortality.
`Australian studies show that most heroin overdose
`cases are attended by ambulance [19,23]. Ambulance
`service records are a rich source of information on the
`nature and prevalence of heroin overdose [24–26]. In
`this study, we linked a database of ambulance service
`records with a mortality database to examine trends in
`overdose mortality among people who had previously
`experienced a non-fatal heroin overdose. This retro-
`spective cohort study design allowed us to examine
`overdose mortality among a cohort who had been pre-
`viously attended by ambulance for a non-fatal heroin
`overdose between 2000 and 2005, as well as explore the
`effects of personal characteristics, such as age, sex and
`number of recorded overdoses on mortality rates.
`
`Methods
`
`Case ascertainment
`
`Cases were obtained from a database of ambulance
`service records collated by Turning Point Alcohol and
`Drug Centre in collaboration with the Melbourne Met-
`ropolitan Ambulance Service (MAS; see [27]). This
`database is a compilation of patient care records (PCR)
`that paramedics complete at the scene of attendance.
`PCR include descriptions of attendance and outcome
`details, basic patient demographics and relevant clinical
`details. All cases related to drug overdose or poisoning
`PCR are extracted by trained sorters from the entire
`pool of PCR and sent for data entry by trained coders.
`Although it is difficult to define the term heroin over-
`dose adequately [3], the MAS database allows for a
`reasonably precise definition of overdose. To this end a
`heroin overdose was defined as a case where: (i) a
`positive response (increased respiration rate or Glasgow
`Coma Score) to the administration of naloxone was
`observed and there was no indication that the overdose
`resulted from another opioid, such as codeine or
`methadone; or (ii) heroin use is established through the
`assessment of the ambulance paramedic or by another
`person at the scene but naloxone was not administered.
`A more detailed description of the MAS drug overdose
`and poisoning database is reported elsewhere [27]. All
`persons recorded as experiencing a non-fatal heroin
`overdose attended by ambulance between 1 October
`2000 and 31 September 2005 were included in the
`cohort.
`
`Data matching
`
`At data entry each case on the MAS database is given
`an alpha-numeric personal identifier code made up a
`
`© 2009 Australasian Professional Society on Alcohol and other Drugs
`
`letters from patient’s first and second name, their date
`of birth and sex. The codes for our study cohort were
`sent to the National Death Index (NDI) for retrospec-
`tive matching. The NDI is a data register containing
`information on all deaths occurring in Australia since
`1980. The NDI is designed to facilitate the conduct of
`epidemiological studies and its use is strictly confined
`to medical research. The NDI matched personal iden-
`tifier codes to all mortality cases recorded between
`2000 and 2005. Matched cases were identified and date
`of death and International Classification of the Diseases
`(ICD) 10 cause of death codes were merged with the
`MAS dataset. Overdose deaths were defined as cases
`where accidental poisoning (ICD-10 X40-X49), inten-
`tional self-poisoning (ICD-10 X60-X64) and/or poi-
`soning by drugs/toxic substances (ICD-10 T36-T65)
`were assigned as primary or secondary causes of death.
`Taking account of the polydrug use norms among
`heroin injectors and the imprecise nature of death reg-
`istrations and identification of causes of death (particu-
`larly in drug-related mortality), a broad range of drug
`classes was included to ascertain drug overdose mor-
`tality among heroin injectors.
`
`Analysis
`
`Overdose mortality was estimated for all cases, with the
`person-years (PY) of follow-up determined from date
`of first overdose attendance to either death or finally
`censored at 31 December 2005 (the last date at which
`full mortality data were available at the time of data
`matching). Trends in mortality were described for the
`entire cohort and by sex, age at baseline and number of
`previous overdose events attended by ambulance
`between 1 October 2000 and 31 September 2005.
`Comparisons of mortality among these groups were
`analysed using Kaplan–Meier survival analyses (for
`baseline characteristics) and a continuous time Cox
`regression analysis. Non-fatal overdose events were
`treated as a continuously time-varying covariate, with
`follow-up time divided into periods between each non-
`fatal overdose event for each participant. All analyses
`were conducted using Intercooled stata version 9.0
`(StataCorp, College Station, TX, USA).
`
`Results
`
`A total of 4884 people were followed from the date of
`their first non-fatal heroin overdose ambulance atten-
`dance recorded after 1 October 2000 until the date of
`death or 31 December 2005, an average of 2.24 years
`between overdose events or censorship (minimum
`0.003 years, maximum 5.25 years). The mean age of
`cases at ascertainment was 29 years (SD 8.28) with
`1627 (33%) aged less than 25 years, 1238 (25%) aged
`
`Opiant Exhibit 2023
`Nalox-1 Pharmaceuticals, LLC v. Opiant Pharmaceuticals, Inc.
`IPR2019-00690
`Page 2
`
`

`

`Overdose mortality following heroin overdose
`
`349
`
`Table 1. Mortality rate of injecting drug users that had previously experienced non-fatal overdose
`
`Number in
`cohorta
`
`Cumulative non-fatal
`heroin overdoses
`
`Overdose
`deaths (%)
`
`Person-years
`
`Overall
`
`95% confidence interval
`
`Overdose mortality rate
`per 100 person-years
`
`Annual rates
`2000
`2001
`2002
`2003
`2004
`2005
`Total rate
`Sex
`Male
`Female
`Age group (years)
`<25
`25–29
`30–34
`35+
`
`868
`1 619
`2 357
`3 321
`4 223
`4 699
`4 884
`
`3 441
`1 443
`
`1 627
`1 238
`871
`1 150
`
`989
`1 875
`2 833
`4 172
`5 552
`6 445
`
`4 531
`1 914
`
`2 205
`1 657
`1 137
`1 446
`
`10 (6)
`15 (9)
`22 (13)
`33 (20)
`46 (28)
`38 (23)
`164
`
`131 (80)
`33 (20)
`
`43 (26)
`51 (31)
`22 (13)
`48 (29)
`
`104
`1 280
`2 030
`2 639
`3 717
`4 427
`13 629
`
`9 463
`4 167
`
`5 107
`3 348
`2 243
`2 932
`
`9.65
`1.17
`1.03
`1.25
`1.24
`0.86
`1.20
`
`1.38
`0.79
`
`0.84
`1.52
`0.98
`1.64
`
`5.19–17.93
`0.71–1.94
`0.67–1.59
`0.89–1.76
`0.93–1.65
`0.63–1.18
`1.03–1.40
`
`1.17–1.64
`0.56–1.11
`
`0.63–1.14
`1.16–2.01
`0.65–1.49
`1.23–2.17
`
`aParticipant numbers for annual rates denote sample size at the end of each calendar year.
`
`between 25 and 29 years, 871 (18%) aged 30–34 years
`and 1150 (24%) aged 35 years and over. The majority
`(71%) of non-fatal overdose cases were male.
`A total of 164 people died of an overdose during the
`follow-up period. Of these, 131 (80%) were male, 101
`(62%) had been previously attended to by an ambulance
`for one heroin overdose, 36 (22%) had twice been
`attended to for a previous heroin overdose and 27 (17%)
`had been attended to more than twice for previous
`heroin overdoses (maximum six) during the follow-up
`period. The overall overdose mortality rate was 1.20
`(95% CI, 1.03–1.40) per 100 PY. However, this varied
`dramatically across the follow-up period; from 9.65 per
`100 PY in 2000 to around 1 per 100 PY for subsequent
`years. Mortality rates were higher for males and those
`aged 25–29 years and over 35 years (Table 1).
`Differences in survival were assessed using Kaplan–
`Meier survival curves and Cox proportional hazards
`modelling. Figure 1 shows Kaplan–Meier curves by the
`baseline characteristics of sex and age. Table 2 shows
`unadjusted and adjusted hazard ratios by sex, age and
`number of previous non-fatal overdoses experienced
`during the follow-up period. Being male, aged 25–29
`and over 35 years and having been attended by ambu-
`lance multiple times for previous non-fatal heroin over-
`doses was associated with greater risk of overdose
`mortality.
`
`Discussion
`
`This is the first study to provide a detailed examination
`of overdose mortality among a cohort who had previ-
`
`ously been attended by ambulance for a non-fatal
`heroin overdose. During a follow-up period of 5 years
`and 3 months, 4884 individuals were identified (cumu-
`lative PY of 13 629) accounting for 6445 cases of non-
`fatal heroin overdose ambulance attendances and 164
`overdose deaths. Our data showed a ratio of non-fatal
`to fatal overdose of 39.3:1 or that 2.6% of overdoses
`result in mortality—findings consistent with previously
`reported Australian data (3.1%, see [14]) and anec-
`dotal reports from Australian heroin users (5%, see
`[23]).
`The overall annual rate of mortality was 1.20 per 100
`PY, but fluctuated approximately 1 per 100 PY across
`most of the follow-up period. Overdose mortality rates
`reported elsewhere vary across time, location and other
`factors, such as recruitment sites (e.g. treatment vs.
`emergency department) and prevalence of HIV among
`IDU. However, in all years other than 2000 our esti-
`mates of rates of overdose mortality are largely consis-
`tent with other Australian and international studies of
`heroin-related mortality [2,7,8,28,29]. In light of the
`case ascertainment used in this study and findings pre-
`sented here and in other research of an increased over-
`dose risk associated with previous overdose events (e.g.
`[17]), we might have expected rates of mortality in this
`study to be comparably higher than those reported in
`other studies.
`Our estimated annual mortality for 2000 was 9.65
`per 100 PY, some 9–10 times higher
`than that
`observed in subsequent years. This figure stresses the
`effects of the heroin ‘glut’ [30] of the late 1990s and
`the subsequent heroin ‘drought’ [31]. The glut condi-
`
`© 2009 Australasian Professional Society on Alcohol and other Drugs
`
`Opiant Exhibit 2023
`Nalox-1 Pharmaceuticals, LLC v. Opiant Pharmaceuticals, Inc.
`IPR2019-00690
`Page 3
`
`

`

`350
`
`M. A. Stoové et al.
`
`(a)
`
`(b)
`
`Figure 1. Kaplan–Meier survival curves of overdose mortality
`by (a) sex and (b) age at cohort entry.
`
`tions of heroin supply of high purity and ready avail-
`ability of the late 1990s appear to have driven the
`extraordinarily high rates of overdose mortality we
`observed in 2000. The effect of the heroin drought,
`where indicators of heroin purity and availability fell
`sharply in Melbourne, was to reduce rates of heroin
`overdose mortality and morbidity by approximately
`80% [31]. These effects were evident in our cohort
`with mortality rates among our sample more consis-
`tent (even somewhat less than) with those previously
`reported, perhaps reflecting more ‘normal’ heroin
`market conditions in Melbourne following the onset
`of the heroin drought [30]. The effects are important
`in an international context as the heroin drought was
`not unique to Australia with researchers in western
`Canada also reporting similar market disruptions that
`produced comparable, if less dramatic, reductions in
`
`© 2009 Australasian Professional Society on Alcohol and other Drugs
`
`heroin overdose mortality [32]. This considerable fluc-
`tuation in mortality also has implications for indirect
`prevalence estimates of IDU populations, given that
`mortality multipliers remain a popular method to esti-
`mate the size of this ‘hidden’ population.
`findings
`Consistent with
`previous
`research
`[2,4,7,26,28], males were over-represented in both
`non-fatal and fatal overdose cases, and showed an
`increased mortality risk. Also consistent with previous
`findings was the association between older age and
`increased risk of fatal overdose. Although contrary to
`popular belief, it is well documented elsewhere that
`fatal heroin overdose is not concentrated among
`younger, novice or inexperienced users. Rather over-
`dose mortality occurs on average among those in their
`late twenties and early thirties with generally consider-
`able experience of heroin use [1,8,33].
`Our study has shown, for the first time, the profound
`effect of previous non-fatal overdoses on subsequent
`risk of overdose fatality. After controlling for age and
`sex, those who had been attended by an ambulance for
`two non-fatal overdoses in the follow-up period were at
`more than three and half times the risk of a fatal over-
`dose compared with those who had experienced only
`one overdose in this period. This figure climbed alarm-
`ingly to more than seven times the risk for those who
`had experienced more than two non-fatal overdoses.
`This result suggests that the risk of overdose among
`heroin users is not attenuated through knowledge
`obtained from previous overdose experiences. Indeed,
`overdose risk appears to increase considerably follow-
`ing previous events.
`These results have implications for overdose preven-
`tion. Overdose prevention initiatives are numerous,
`mostly focussing on education programs of factors
`associated with overdose risk [34]. Such programs have
`attracted criticism because they fail to acknowledge
`contextual factors associated with overdose that impede
`the uptake of risk reduction advice, such as not inject-
`ing alone or avoiding polydrug use [26,35]. Calls for
`structural interventions that control risk environments
`for IDU, such as the introduction of supervised injec-
`tion facilities (SIF) are motivated in part by the aim to
`reduce overdose incidence. Indeed, studies of SIF
`suggest that they are effective in managing overdose
`and reducing the effect of overdose on drug-related
`mortality and morbidity [22]. Furthermore, in findings
`similar to ours, a study of North America’s first SIF in
`Vancouver showed that having a history overdose was a
`significant independent predictor of time to overdose
`among the 336 on-site overdoses observed [6].The fact
`that none of these overdoses in Vancouver was fatal
`suggests the importance of SIF in reducing overdose
`fatality and creating environments protective against
`overdose mortality. Indeed, our results suggest that
`
`Opiant Exhibit 2023
`Nalox-1 Pharmaceuticals, LLC v. Opiant Pharmaceuticals, Inc.
`IPR2019-00690
`Page 4
`
`

`

`Overdose mortality following heroin overdose
`
`351
`
`Table 2. Cox proportional hazard ratios: unadjusted and adjusteda analyses
`
`Unadjusted
`
`Adjusted
`
`Hazard ratio
`
`95% confidence interval
`
`Hazard ratio
`
`95% confidence interval
`
`Sex
`Female
`Male
`Age group (years)
`<25
`25–29
`30–34
`35+
`Number of previous overdoses
`1
`2
`>2
`
`aAdjusted for variables in the Table.
`
`1
`1.70
`
`1
`1.68
`1.05
`1.75
`
`1
`3.60
`7.20
`
`those with a history of non-fatal overdose would benefit
`most from the use of such facilities.
`There are a number of limitations associated with
`this study primarily related to cohort ascertainment.
`First, the cohort was derived from records of ambu-
`lance attendances at overdose events between 2000 and
`2005. Although we believe that most overdose events in
`Melbourne are attended by ambulance, and heroin
`users are not reluctant to call ambulances in this juris-
`diction [19,36,37], it is possible that some cases had
`also experienced overdoses that were not attended by
`an ambulance. However, overdose events attended by
`ambulance are likely to identify more severe non-fatal
`overdose events and perhaps indicative of individuals
`using heroin in particularly risky ways. In the context of
`survival, it is this group of injectors that we might
`reason to be of greatest interest. It is also of interest
`that, despite experiencing multiple overdoses of suffi-
`cient severity to require ambulance attendance, lessons
`are perhaps not being learned from these events to
`minimise later overdose mortality risk. Second, the
`finite period of follow-up means that overdose events
`occurring before October 2000 could not be deter-
`mined. Before October 2000, case identifiers contained
`only the five letter alpha-code and year of birth (con-
`sidered insufficient to reliably match data with the
`NDI). Both these limitations have particular implica-
`tions for the associations between cumulative overdose
`experiences and subsequent overdose mortality. Third,
`approximately 10% of cases did not or could not
`provide ambulance paramedics with personal identify-
`ing information, thus limiting the number of individu-
`als included in the dataset or limiting the number of
`overdoses attributed to individuals included in the
`analyses. It is plausible that this group might be most at
`
`—
`1.16–2.50
`
`—
`1.12–2.53
`0.63–1.76
`1.16–2.64
`
`—
`2.45–5.26
`4.69–11.04
`
`1
`1.67
`
`1
`1.63
`1.03
`1.79
`
`1
`3.71
`7.38
`
`—
`1.14–2.45
`
`—
`1.09–2.45
`0.62–1.73
`1.19–2.72
`
`—
`2.54–5.45
`4.81–11.32
`
`risk of overdose if the severity of their overdose event
`was such that they were unable to provide identifying
`information with sufficient clarity. Finally, the classifi-
`cation of a heroin overdose is difficult and some mis-
`classification might have occurred in ascertaining cases
`for this study.
`
`Conclusions
`
`Independent of sex and age, the results of this studied
`showed for the first time that experiencing a non-fatal
`overdose substantially increases the risk of subsequent
`overdose mortality. Annual overdose mortality reported
`in this study corresponded well with known changes in
`heroin market dynamics. Following a period of high
`heroin purity and availability and high mortality rates,
`overdose mortality stabilised at levels comparable with
`those reported in other studies in Australia and inter-
`nationally. These findings provide salient evidence of
`personal and temporal
`factors that contribute to
`increased risk of fatal overdose events. Strategies to
`effectively reduce overdose mortality should consider
`personal drug use histories and drug market dynamics
`as important factors in determining overdose risk.
`
`Acknowledgements
`
`The authors would like to thank John Harding, Mark
`Short and Kun Zhao from the Australian Institute of
`Health and Welfare for their advice and assistance in
`facilitating the data linkage component of this study.We
`would like to acknowledge Stefan Cvetkovski for his
`assistance in preparing and providing ambulance over-
`dose attendance data and Ella Bouchard for her advice
`in the preparation of this manuscript. We would also
`
`© 2009 Australasian Professional Society on Alcohol and other Drugs
`
`Opiant Exhibit 2023
`Nalox-1 Pharmaceuticals, LLC v. Opiant Pharmaceuticals, Inc.
`IPR2019-00690
`Page 5
`
`

`

`352
`
`M. A. Stoové et al.
`
`like to thank Turning Point Alcohol and Drug Centre
`for providing infrastructure support to undertake this
`study.This study was approved by the Human Research
`Ethics Committee of the Department of Human Ser-
`vices, State Government of Victoria.
`
`References
`[1] Warner-Smith M, Darke S, Lynskey M, Hall W. Heroin
`overdose: causes and consequences. Addiction 2001;
`96:1113–25.
`[2] Hickman M, Carnwath Z, Madden P, et al. Drug-related
`mortality and fatal overdose risk: pilot cohort study of
`heroin users recruited from specialist drug treatment sites in
`London. J Urban Health 2003;80:274–87.
`[3] Darke S, Zador D. Fatal heroin ‘overdose’: a review. Addic-
`tion 1996;91:1765–72.
`[4] Preti A, Miotto P, De Coppi M. Deaths by unitentional
`illicit overdose in Italy, 1984–2000. Drug Alcohol Depend
`2002;66:275–82.
`[5] Darke S, Degenhardt L, Mattick R. Mortality among illicit
`drug users: epidemiology, causes and intervention. Cam-
`bridge: Cambridge University Press, 2006.
`[6] Kerr T, Tyndall M, Lai C, Montaner J, Wood E. Drug-
`related overdoses within a medically supervised safer injec-
`tion facility. Int J Drug Policy 2006;17:436–41.
`[7] Bartu A, Freeman NC, Gawthorne GS, Codde JP, Holman
`DJ. Mortality in a cohort of opiate and amphetamine users
`in Perth, Western Australia. Addiction 2004;99:53–60.
`[8] Gossop M, Stewart D, Treacy S, Marsden J. A prospective
`study of mortality among drug misusers during a 4-year
`period after seeking treatment. Addiction 2002;97:39–47.
`[9] Engstrom A, Adamsson C, Allbeck P, Rydberg U. Mortality
`in patients with substance abuse: a follow-up in Stockholm
`county, 1973–1984. Int J Addict 1991;26:91–106.
`[10] Sanchez-Carbonell X, Seus L. Ten-year survival analysis of
`a cohort of heroin addicts in Catalonia: the EMETYST
`project. Addiction 2000;95:941–8.
`[11] Vlahov D, Wang C, Galai N, et al. Mortality risk among
`new onset injection drug users. Addiction 2004;99:946–54.
`[12] Capelhorn JRM. Deaths in the first two weeks of mainte-
`nance treatment in NSW in 1994: identifying cases of iatro-
`genic methadone toxicity. Drug Alcohol Rev 1998;17:9–17.
`[13] Capelhorn JRM, Drummer OH. Mortality associated with
`New South Wales methadone programs in 1994: lives lost
`and saved. Med J Aust 1999;170:104–9.
`[14] Darke S, Mattick R, Degenhardt L. The ratio of non-fatal to
`fatal heroin overdose. Addiction 2003;98:1169–72.
`[15] Brugal MT, Barrio G, De LF, Regidor E, Royuela L,
`Suelves JM. Factors associated with non-fatal heroin over-
`dose: assessing the effect of frequency and route of heroin
`administration. Addiction 2002;97:319–27.
`[16] Fischer B, Brissette S, Brochu S, et al. Determinants of
`overdose incidents among illicit opioid users in five Cana-
`dian cities. Can Med Assoc J 2004;171:235–9.
`[17] Kerr T, Fairbairn N, Tyndall M, et al. Predictors of non-
`fatal overdose among a cohort of polysubstance-using injec-
`tion drug users. Drug Alcohol Depend 2007;87:39–45.
`[18] Ochoa KC, Davidson PJ, Evans JL, Hahn JA, Page-Shafer
`K, Moss AR. Heroin overdose among young injection drug
`users in San Francisco. Drug Alcohol Depend 2005;
`80:297–302.
`[19] Warner-Smith M, Darke S, Day C. Morbidity associated
`with non-fatal heroin overdose. Addiction 2002;97:963–7.
`
`© 2009 Australasian Professional Society on Alcohol and other Drugs
`
`[20] Darke S, Hall W. Heroin overdose: research and evidence-
`based intervention. J Urban Health 2003;80:189–200.
`[21] Darke S, Williamson A, Ross J, Mills KL, Havard A,
`Teesson M. Patterns of non-fatal heroin overdose over a
`3-year period: findings from the Australian Treatment
`Outcome Study. J Urban Health 2007;84:283–91.
`[22] van Beek I, Dakin A, Kimber J, Gilmour S. The Sydney
`supervised injecting centre: reducing harm associated with
`heroin overdose. Crit Public Health 2004;14:391–406.
`[23] Darke S, Ross J, Hall W. Overdose among heroin users in
`Sydney, Australia: I. Prevalence and correlates of non-fatal
`overdose. Addiction 1996;91:405–11.
`[24] Degenhardt L, Hall W, Adelstein BA. Ambulance calls to
`suspected drug overdoses: New South Wales patterns July
`1997 to June 1999. Aust N Z JPublic Health 2001;25:447–
`50.
`[25] Dietze P, Jolley D, Cvetkovski S. Patterns and characteris-
`tics of ambulance attendance at heroin overdose at a local
`area level in Melbourne: implications for service provision.
`J Urban Health 2003;80:248–60.
`[26] Dietze P, Jolley D, Fry C, Bammer G. Transient changes in
`behaviour lead to heroin overdose: results from a case-
`crossover study of non-fatal overdose. Addiction 2005;
`100:636–42.
`[27] Dietze P, Cvetkovski S, Rumbold G, Miller P. Ambulance
`attendance at heroin overdose in Melbourne: the establish-
`ment of a database of ambulance service records. Drug
`Alcohol Rev 2000;19:27–33.
`[28] Brugal MT, Domingo-Salvany A, Puig R, Barrio G, Garcia
`de Olalla P, de la Fuente L. Evaluating the impact of metha-
`done maintenance programmes on mortality due to over-
`dose and AIDS in a cohort of heroin users in Spain.
`Addiction 2005;100:981–9.
`[29] Caplehorn JR, Dalton MS, Haldar F, Petrenas AM, Nisbet
`JG. Methadone maintenance and addicts’ risk of
`fatal
`heroin overdose. Subst Use Misuse 1996;31:177–96.
`[30] Dietze P, Fitzgerald J. Interpreting changes in heroin supply
`in Melbourne: droughts gluts or cycles? Drug Alcohol Rev
`2002;21:295–303.
`[31] Degenhardt L, Day C, Dietze P, et al. Effects of a sustained
`heroin shortage in three Australian States. Addiction
`2005;100:908–20.
`[32] Wood E, Stoltz J-A, Li K, Montaner J, Kerr T. Changes in
`Canadian heroin supply coinciding with the Australian
`heroin shortage. Addiction 2006;101:689–95.
`[33] Bargagli AM, Sperati A, Davoli M, Forastiere F, Perucci
`CA. Mortality among problem drug users in Rome: an
`18-year
`follow-up study, 1980–1997. Addiction 2001;
`96:1455–63.
`[34] Seal KH, Downing M, Kral AH, et al. Attitudes about
`prescribing take-home Naloxone to injecting drug users for
`the management of heroin overdose: a survey of street-
`recruited injectors in the San Francisco Bay Area. J Urban
`Health 2003;80:291–301.
`[35] Moore D. Governing street-based injecting drug users: a
`critique of heroin overdose prevention in Australia. Soc Sci
`Med 2004;59:1547–57.
`[36] Kerr D, Dietze P, Jolley D, Kelly A-M. Peer resuscitation
`after witnessed heroin overdose in Melbourne. Drug
`Alcohol Rev 2009; in press.
`[37] Dietze P, Fry C, Rumbold G, Gerostamoulos J. The
`context, management and prevention of heroin overdose in
`Victoria: the promise of a diverse approach. Addict Res
`Theory 2001;9:437–58.
`
`Opiant Exhibit 2023
`Nalox-1 Pharmaceuticals, LLC v. Opiant Pharmaceuticals, Inc.
`IPR2019-00690
`Page 6
`
`