Heroin-related deaths in South Western Sydney, Australia, 1992-96

Concurrent Alcohol and Opioid Use Among Harm Reduction Clients

BACKGROUND

Harm reduction services infrequently address alcohol use among clients using opioids, despite the evaluated risk of overdose or medical consequences for clients with viral infections. The purpose of this study is to assess concurrent alcohol and opioid use among syringe services and overdose prevention program participants predominately in southern Ohio and northern Kentucky.

METHODS

This is a cross-sectional study using self-report data (n = 1,142) pooled across regional overdose prevention programs and a mobile syringe services program. The outcome variable was concurrent use categorized as no concurrent alcohol, prescription opioid or heroin use; alcohol and heroin or prescription opioid use; and alcohol, prescription opioid and heroin use in the past three months.

RESULTS

The sample was predominantly white (95%), 56% were male and the mean age was 33 years old. Forty-seven percent of the clients had no concurrent use of alcohol and opioids; 20.1% reported concurrent use of alcohol and either heroin or prescription opioids; and 33.4% reported concurrent use of alcohol, heroin and prescription opioids in the past 3 months. Lifetime suicidal ideation and non-opioid drug use were associated with concurrent alcohol and opioid use in the multivariable model.

CONCLUSION

Harm reduction clients with concurrent alcohol and opioid use may warrant enhanced overdose prevention services. Syringe services and overdose prevention program participants may benefit from education or a brief intervention on alcohol consumption.

Using the Advocacy Coalition Framework and Multiple Streams policy theories to examine the role of evidence, research and other types of knowledge in drug policy

BACKGROUND AND AIMS

The prevailing 'evidence-based policy' paradigm emphasizes a technical-rational relationship between alcohol and drug research evidence and subsequent policy action. However, policy process theories do not start with this premise, and hence provide an opportunity to consider anew the ways in which evidence, research and other types of knowledge impact upon policy. This paper presents a case study, the police deployment of drug detection dogs, to highlight how two prominent policy theories [the Advocacy Coalition Framework (ACF) and the Multiple Streams (MS) approach] explicate the relationship between evidence and policy.

METHODS

The two theories were interrogated with reference to their descriptions and framings of evidence, research and other types of knowledge. The case study methodology was employed to extract data concerned with evidence and other types of knowledge from a previous detailed historical account and analysis of drug detection dogs in one Australian state (New South Wales). Different types of knowledge employed across the case study were identified and coded, and then analysed with reference to each theory. A detailed analysis of one key 'evidence event' within the case study was also undertaken.

RESULTS

Five types of knowledge were apparent in the case study: quantitative program data; practitioner knowledge; legal knowledge; academic research; and lay knowledge. The ACF highlights how these various types of knowledge are only influential inasmuch as they provide the opportunity to alter the beliefs of decision-makers. The MS highlights how multiple types of knowledge may or may not form part of the strategy of policy entrepreneurs to forge the confluence of problems, solutions and politics.

CONCLUSIONS

Neither the Advocacy Coalition Framework nor the Multiple Streams approach presents an uncomplicated linear relationship between evidence and policy action, nor do they preference any one type of knowledge. The implications for research and practice include the contestation of evidence through beliefs (Advocacy Coalition Framework), the importance of venues for debate (Advocacy Coalition Framework), the way in which data and indicators are transformed into problem specification (Multiple Streams) and the importance of the policy ('alternatives') stream (Multiple Streams).

Development of a proto-typology of opiate overdose onset

BACKGROUND AND AIMS

The time available to act is a crucial factor affecting the probable success of interventions to manage opiate overdose. We analyse opiate users' accounts of non-fatal overdose incidents to (i) construct a proto-typology of non-fatal opiate overdose onset and (ii) assess the implications for overdose management and prevention of fatalities.

METHODS

Re-analysis of a subset of data from a large qualitative study of non-fatal opiate overdose conducted from 1997 to 1999. Data were generated from semi-structured interviews undertaken with opiate users who had experienced a non-fatal overdose in the previous 24 hours. Forty-four participants (30 men; 14 women; aged 16-47 years) provided sufficient information for in-depth analysis. Data relating to 'memory of the moment of overdose', 'time to loss of consciousness' and 'subjective description of the overdose experience' were scrutinised using iterative categorization.

FINDINGS

Four types of overdose onset were identified: type A 'amnesic' (n = 8), characterized by no memory, rapid loss of consciousness and no description of the overdose experience; type B 'conscious' (n = 17), characterized by some memory, sustained consciousness and a description of the overdose in terms of feeling unwell and symptomatic; type C 'instant' (n = 14), characterized by some memory, immediate loss of consciousness and no description of the overdose experience; and type D 'enjoyable' (n = 5), characterized by some memory, rapid loss of consciousness and a description of the overdose experience as pleasant or positive.

CONCLUSIONS

The identification of different types of overdose onset highlights the complexity of overdose events, the need for a range of interventions and the challenges faced in managing incidents and preventing fatalities. Opiate overdose victims who retain consciousness for a sustained period and recognize the negative signs and symptoms of overdosing could summon help or self-administer naloxone, thus indicating that opiate overdose training should incorporate self-management strategies.

The Sydney Medically Supervised Injecting Centre: Client Characteristics and Predictors of Frequent Attendance during the First 12 Months of Operation

This paper describes characteristics of clients registered in the first 12 months of the Sydney Medically Supervised Injecting Centre's (MSIC) operation, as well as predictors of frequent attendance. The study is based on information collected from clients at their initial registration and subsequent service utilization. Most of the 2,719 clients were male (71%), almost half had previously experienced at least one nonfatal heroin overdose, and one quarter had accessed formal drug treatment in the previous 12 months. Characteristics associated with frequent attendance at the MSIC were reporting previous attendance at the local primary health service for injection drug users (IDU), injecting drugs other than amphetamine, reporting sex work, injecting at least daily, and injecting in a public place in the month before registration.

Ethanol reversal of cellular tolerance to morphine in rat locus coeruleus neurons

Consumption of ethanol is a considerable risk factor for death in heroin overdose. We sought to determine whether a mildly intoxicating concentration of ethanol could alter morphine tolerance at the cellular level. In rat locus coeruleus (LC) neurons, tolerance to morphine was reversed by acute exposure of the brain slice to ethanol (20 mM). Tolerance to the opioid peptide [d-Ala(2),N-MePhe(4),Gly-ol]-enkephalin was not reversed by ethanol. Previous studies in LC neurons have revealed a role for protein kinase C (PKC)α in μ-opioid receptor (MOPr) desensitization by morphine and in the induction and maintenance of morphine tolerance, but we have been unable to demonstrate that 20 mM ethanol produces significant inhibition of PKCα. The ability of ethanol to reverse cellular tolerance to morphine in LC neurons was absent in the presence of the phosphatase inhibitor okadaic acid, indicating that dephosphorylation is involved. In human embryonic kidney 293 cells expressing the MOPr, ethanol reduced the level of MOPr phosphorylation induced by morphine. Ethanol reversal of tolerance did not appear to result from a direct effect on MOPr since acute exposure to ethanol (20 mM) did not modify the affinity of binding of morphine to the MOPr or the efficacy of morphine for G-protein activation as measured by guanosine 5'-O-(3-[(35)S]thio)triphosphate binding. Similarly, ethanol did not affect MOPr trafficking. We conclude that acute exposure to ethanol enhances the effects of morphine by reversing the processes underlying morphine cellular tolerance.

Phenazepam abuse in the UK: an emerging problem causing serious adverse health problems, including death

OBJECTIVE

Phenazepam (fenazepam; 7-bromo-5-(2-chlorophenyl)-1,3-dihydro-2H-1,4-benzodiazepin-2-one; PNZ, 'Bonsai') is a benzodiazepine developed in the former Soviet Union during the 1970s to treat neurological disorders, epilepsy, and alcohol withdrawal syndrome. Its recreational use appears to have increased over recent years. Because of the lack of accessible data on this substance, it is important that information is made available to health professionals.

METHODS

A literature search was conducted in relevant databases (Medline, Toxbase, PsychInfo, etc.), grey literature (using Google Scholar) and Internet sites to identify key data on phenazepam, including epidemiology such as availability, price, supply sources, confiscations, and health-related problems.

RESULTS

Information from these sources indicates the potential for serious adverse health consequences for this drug when taken recreationally and that its use is spreading in the USA and Europe. Although first use was reported in the UK in October 2009, major concerns in the UK arose in summer 2010 when individuals across Britain were admitted to hospital following overdose. Nine UK fatalities were reported in which phenazepam was detected in post mortem toxicology but not implicated in death. The first UK death directly involving phenazepam was notified in July and the second in November 2011.

CONCLUSIONS

This paper summarises the key information about phenazepam abuse and health problems of which health professionals, especially those in Emergency Departments, should be aware and presents new information in respect of fatalities caused by the drug.

Understanding drug-related mortality in released prisoners: a review of national coronial records

Background

The prisoner population is characterised by a high burden of disease and social disadvantage, and ex-prisoners are at increased risk of death following release. Much of the excess mortality can be attributed to an increased risk of unnatural death, particularly from drug overdose; however, relatively few studies have investigated the circumstances surrounding drug-related deaths among released prisoners. This study aimed to explore and compare the circumstances of death for those who died from accidental drug-related causes to those who died from all other reportable causes.

Methods

A nationwide search of the Australian National Coroners Information System (NCIS) was conducted to identify reportable deaths among ex-prisoners from 2000 to 2007. Using a structured coding form, NCIS records for these cases were interrogated to explore causes and circumstances of death.

Results

Coronial records for 388 deceased ex-prisoners were identified. Almost half of these deaths were a result of accidental drug-related causes (45%). The majority of accidental drug-related deaths occurred in a home environment, and poly-substance use at or around the time of death was common, recorded in 72% of drug-related deaths. Ex-prisoners who died of accidental drug-related causes were on average younger and less likely to be Indigenous, born in Australia, married, or living alone at or around the time of death, compared with those who died from all other reportable causes. Evidence of mental illness or self-harm was less common among accidental drug-related deaths, whereas evidence of previous drug overdose, injecting drug use, history of heroin use and history of drug withdrawal in the previous six months were more common.

Conclusions

Drug-related deaths are common among ex-prisoners and often occur in a home (vs. public) setting. They are often associated with use of multiple substances at or around the time of death, risky drug-use patterns, and even among this markedly disadvantaged group, extreme social disadvantage. These findings reflect the complex challenges facing prisoners upon release from custody and indicate a need to consider drug overdose within the wider framework of ex-prisoner experiences, so that preventive programmes can be appropriately structured and targeted.

Exploring the relationship between homelessness and risk factors for heroin-related death-a qualitative study

The aim of this study was to explore the relationship between housing status, associated social networks and risk factors for heroin-related death. We used semi-structured face-to-face qualitative interviews, recorded, transcribed and analysed thematically by framework techniques at three centres providing services to homeless people in a large cosmopolitan city. Different types of accommodation for homeless people have differing social cultures which have an impact upon the amount of heroin used, likelihood of injecting alone or likelihood of achieving abstinence. Hostel accommodation appeared to be linked with a culture of group injecting, which tends to increase the amount of heroin taken. Those with experience of rough sleeping described heroin use to ameliorate the uncomfortable realities of outdoor sleeping, although the overall amount used tended to be less due to having less money to spend on drugs. The prison setting was described as a setting where heroin use was reduced or stopped. Moving away from homelessness towards sustaining an independent tenancy appeared to be associated with a move towards solitary use. We postulate that a progression towards solitary use in a housed environment is one explanation for previous research findings showing the average age of heroin-related death to be increasing despite a decrease in the average age of initiation into heroin use. Hostel accommodation should form a priority setting for future health promotion interventions aimed to reduce heroin-related death. They appear to be linked with an increase in heroin use in the presence of a third party. Drug users sleeping rough in cold climates need to be made aware of the dangers of medicating with heroin to address problems of insomnia due to cold weather.

A systematic review of research examining benzodiazepine-related mortality

PURPOSE

This paper will review literature examining the association of benzodiazepine use and mortality.

METHODS

An extensive literature review was undertaken to locate all English-language published articles that examine mortality risk associated with use of benzodiazepines from 1990 onwards.

RESULTS

Six cohort studies meeting the criteria above were identified. The results were mixed. Three of the studies assessed elderly populations and did not find an increased risk of death associated with benzodiazepine use, whereas another study of the general population did find an increased risk, particularly for older age groups. A study of a middle aged population found that regular benzodiazepine use was associated with an increased mortality risk, and a study of 'drug misusers' found a significant relationship between regular use of non-prescribed benzodiazepines and fatal overdose. Three retrospective population-based registry studies were also identified. The first unveiled a high relative risk (RR) of death due to benzodiazepine poisoning versus other outcomes in patients 60 or older when compared to those under 60. A positive but non-significant association between benzodiazepine use and driver-responsible fatalities in on-road motor vehicle accidents was reported. Drug poisoning deaths in England showed benzodiazepines caused 3.8% of all deaths caused by poisoning from a single drug.

CONCLUSION

On the basis of existing research there is limited data examining independent effects of illicit benzodiazepine use upon mortality. Future research is needed to carefully examine risks of use in accordance with doctors' prescriptions and extra-medical use.

Does alcohol increase the risk of overdose death: the need for a translational approach

BACKGROUND

We argue for a translational approach to addiction science, using an important current research question as a case study.

CASE STUDY

What is the evidence in support of the hypothesis that alcohol increases the risk of a heroin/opiate overdose through a pharmacological interaction?

FINDINGS

The positive epidemiological evidence shows that opiate overdose deaths rarely involve a single drug; that alcohol is the most common other drug involved; that there is a negative association between alcohol and morphine concentration at post mortem; and that post-mortem levels of morphine are often below the levels expected of highly tolerant individuals. The evidence is consistent with the hypothesis that heroin users who drink may require less heroin to overdose than those who do not drink (all other factors being equal) because of a pharmacological interaction. However, the evidence is consistent with, and does not rule out, other causal (and non-causal) pathways. Alcohol could be associated negatively with tolerance, or confounded by other factors. Experimental evidence is required which is unlikely to be obtained through further epidemiological study or through randomized clinical trials.

CONCLUSIONS

We believe that animal models could provide the key evidence to test the hypothesis for a 'pharmacodynamic' or 'pharmacokinetic' interaction, which could be corroborated in clinical challenge studies and epidemiological studies. Such a translational approach demands greater collaboration between addiction scientists from basic to applied science and from neuroscience to social science, and would be able to address other key research questions and hypotheses in addiction.

Sustained-release naltrexone: novel treatment for opioid dependence

The devastating costs of opioid abuse and dependence underscore the need for effective treatments for these disorders. At present, several different maintenance medications exist for treating opioid dependence, including methadone, buprenorphine and naltrexone. Of these, naltrexone is the only one that possesses no opioid agonist effects. Instead, naltrexone occupies opioid receptors and prevents or reverses the effects produced by opioid agonists. Despite its clear pharmacologic effectiveness, its clinical effectiveness in treating opioid dependence has been disappointing, primarily due to non-compliance with taking the medication. However, the recent availability of sustained-release formulations of naltrexone has renewed interest in this medication. The present paper describes the development of sustained-release naltrexone formulations and discusses the clinical issues associated with their use in treating opioid dependence.

Reducing hospital presentations for opioid overdose in patients treated with sustained release naltrexone implants

BACKGROUND

Non-fatal overdoses represent a significant morbidity for regular heroin users. Naltrexone is an opioid antagonist capable of blocking the effects of heroin, thereby preventing accidental overdose. However, treatment with oral naltrexone is often associated with non-compliance. An alternative is the use of a sustained release preparation of naltrexone. The aim of this study was to assess the change in number of opioid and other drug overdoses in a large cohort of heroin dependent persons (n=361; 218 males) before and after treatment with a sustained release naltrexone implant. A sub-group of this cohort (n=146; 83 males) had previously received treatment with oral naltrexone, which also allowed a comparison of overdoses pre- and post-oral and also post-implant treatments.

METHOD

We used a pre-post design, with data prospectively collected via the West Australian Health Services Research Linked Database, and the Emergency Department Information System. Participants were treated under the Australian Therapeutic Goods Administration's special access guidelines.

RESULTS

Most (336, 93%) of the cohort was in one or both databases. We identified 21 opioid overdoses involving 20 persons in the 6 months pre-treatment that required emergency department presentation or hospital admission: none were observed in the 6 months post-treatment. This is consistent with the existing pharmacokinetic data on this implant, which indicates maintenance of blood naltrexone levels at or above 2 ng/ml for approximately 6 months. A reduced number of opioid overdoses were also observed 7-12 months post-implant. The study found a significant increase in sedative "overdoses", some of which occurred in the 10 days following implant treatment and were likely associated with opioid withdrawal and/or implant treatment. For those previously treated with oral naltrexone, more opioid overdoses occurred in both the 6-months prior to and after oral compared to the 6-months post-implant treatment.

CONCLUSIONS

The findings support the clinical efficacy of this sustained release naltrexone implant in preventing opioid overdose. However, given the high prevalence of poly-substance use among dependent heroin users, programs offering this type of treatment should also focus on preventing, detecting and managing poly-substance use.

HIV education in a Siberian prison colony for drug dependent males

Aim

To evaluate the effectiveness of an HIV peer training program conducted in a colony for drug dependent male prisoners in Siberia, Russia.

Method

Questionnaires were used to collect data pre and post peer training sessions. Three peer training sessions were conducted between questionnaires. Fifteen to twenty inmates were trained as peer educators at each week-long health education training session.

Results

In 2000 and 2001, 153 and 124 inmates completed the questionnaire respectively. Respondents in both years reported similar health and injecting histories and comparable levels of sexual activity. Respondents in 2001 were significantly more likely to correctly identify both how HIV can and cannot be transmitted compared to respondents in 2000. The prevalence of tattooing in prison decreased significantly between questionnaires. However, there was virtually no reported use of bleach to clean tattooing or injecting equipment in either 2000 or 2001. Access to condoms increased significantly between questionnaires.

Conclusions

While this training program was associated with improved HIV knowledge, the Ministry of Justice should consider improved and additional harm reduction strategies. These include increased availability of bleach and condoms and the introduction of methadone treatment and syringe exchange in prison.

Mortality in a cohort of opiate and amphetamine users in Perth, Western Australia

AIMS

This study compares the hazard of death among opiate and amphetamine using clients who accessed drug treatment with individuals who had no specialist treatment contact between 1985 and 1998.

DESIGN, SETTING, PARTICIPANTS

This was a retrospective cohort study of 4280 drug-using individuals (2887 opiate users, 1393 amphetamine users) admitted to Perth metropolitan hospitals or Perth psychiatric institutions between 1985 and 1998. Of these, 1469 attended Next Step Specialist Drug and Alcohol Services (928 received methadone and 541 attended counselling or support groups) and 2811 had no contact with this service.

METHODS

Data from two drug treatment programmes were linked with hospital morbidity, psychiatric services and the mortality database using record linkage.

FINDINGS

The results show that people who were currently in drug treatment had a lower hazard of death compared with non-clients and those who had ceased treatment. Those who had ceased treatment more than 6 months ago had 7.0 times the hazard of all-cause death and 8.4 times the hazard of drug-cause death. Opiate users were at 1.4 times the hazard of all-cause death and 2.4 times the hazard of drug-cause death compared with amphetamine users. Males were at 1.79 times the hazard of all-cause death and, unexpectedly, were found to be at 2.69 times the hazard of drug-cause death compared with females.

CONCLUSIONS

Treatment protected clients from premature death compared with people who did not receive treatment and also those who ceased treatment. While amphetamine users had a lower risk of mortality compared with opiate users, the full extent of the relationship between amphetamine use and mortality needs to be examined further.

Peer overdose resuscitation: multiple intervention strategies and time to response by drug users who witness overdose

One hundred and thirty-five drug users in contact with treatment services in Scotland and England were interviewed about their experiences of witnessing overdoses - both overdoses resolved successfully and those leading to death - and actions taken to effect resuscitation. One hundred and four (77%) had witnessed a mean of 11.5 overdoses, of whom 41 (30.4% of the study sample) had witnessed an average of 4.2 fatal overdoses. A wide range of actions was reported at the most recent witnessed overdose, the most common being slapping or shaking the victim (an average of 2.5 minutes after overdose was first recognised) or walking the person around the room (3.2 minutes after recognizing overdose). There was no consistent relationship between the time taken to acting and the number of actions taken. Successful resolution of last witnessed overdose was associated more strongly with immediate onset of overdose, while those that led to death were more often those that involved slow onset of overdose. There is clear evidence of the opportunity and willingness of witnesses to intervene, particularly when overdose onset is immediate, with a wide range of strategies adopted to encourage recovery, although these may often be inappropriate and wrongly prioritized.

Reduced injection risk and sexual risk behaviours after drug misuse treatment: results from the National Treatment Outcome Research Study

This paper investigates injecting, shared use of needles/syringes and sexual risk behaviours at intake to treatment and at one-year follow-up among 753 drug users recruited to the National Treatment Outcome Research Study (NTORS). Injecting, sharing and having unprotected sex were substantially reduced among clients admitted to methadone programmes and among those admitted to residential treatments. The overall levels of risk fell after treatment, and the majority of those who were engaged in high risk behaviours at intake had stopped at follow-up. The results also show the variability of individual outcomes. A minority persisted with their risk behaviour, and others who were not at risk at intake who had started to engage in risky behaviours at follow-up. The behaviour of these clients creates a focal point for risk as well as being a threat to public health. Several social and psychological factors were predictive of health risk behaviours. These included frequency and duration of heroin use, polydrug use, alcohol use, gender, ethnicity, having a drug-using partner, anxiety and depression. The results indicate the important role that can be played by treatment services in helping to reduce the risk of blood-borne infections. We suggest that risk reduction interventions are an important and effective component of treatment programmes.

A prospective study of mortality among drug misusers during a 4-year period after seeking treatment

AIMS

The opportunity to study deaths as they occur within the framework of a prospective cohort study is relatively uncommon. This study investigates deaths among drug misusers over a 4-year period, with specific attention to the circumstances and causes of death, and risk factors for mortality. The study also critically examines the recording of drug-related deaths.

DESIGN, SETTING, PARTICIPANTS

Prospective cohort study of 1075 drug misusers recruited to 54 treatment programmes during 1995.

MEASUREMENTS

Data derived from interviews conducted with clients at intake, death certificates and post-mortem examinations.

FINDINGS

The annual mortality rate was 1.2%, about six times higher than that for a general, age-matched population. Fourteen per cent of the deaths were due to self-inflicted injuries, accidents or violence and 18%, were due to medical causes. The majority of deaths (68%) were associated with drug overdoses. Opiates were the drugs most commonly detected during post-mortem examinations. In the majority of cases, more than one drug was detected. Polydrug use and, specifically, heavy drinking, and use of benzodiazepines and amphetamines, were identified as risk factors for mortality. Anxiety and homelessness were also predictive of increased mortality.

CONCLUSIONS

We suggest that drug misusers and those working with drug misusers need to be more alert to the risks of polydrug use, including the combined use of alcohol with illicit drugs. The study revealed inconsistencies in the recording of drug-related deaths on death certificates. The routine recording of all substances detected during toxicological examination would improve the accuracy of death certification.

Putting harm reduction into an adolescent context

Drug use is now widespread amongst Australian youth. Substance abuse and dependence are becoming increasingly significant health problems. Approximately 50% of 17-year-old Australians report regular consumption of alcohol and nearly 30% report tobacco smoking. The age of onset of substance use is reported to be decreasing. Between 1993 and 1995 the proportion of heroin users who had used the drug before the age of 16 years increased from 2% to 14%. The debate about youth substance use tends to be polarized between the views of Zero Tolerance and Legalization of drugs. The harm reduction approach spans between these two extremes. Examples of harm reduction strategies, such as education campaigns on safe injecting and needle exchange programs, have been effective in curbing the spread of blood-borne viruses such as HIV amongst intravenous drug using youth. The harm reduction approach, taking social context and developmental stage of the individual into account, may also be applied to adolescents at the less extreme end of the substance use spectrum. It is proposed that the harm reduction framework used in this way enables a rational, relevant and consistent response to contemporary youth substance use, aiming to minimize drug related harm.

Heroin-related deaths in New South Wales, Australia, 1992-1996

The coronial files of all heroin-related fatalities that occurred in New South Wales (NSW) over the period 1992-1996 were inspected. There were 953 heroin-related fatalities in NSW over the study period. There was a substantial, statistically significant increase in heroin-related fatalities over the study period, from 152 deaths in 1992 to 226 during 1996. The mean age of cases was 31.0 years, 85% were male, and 85% were classified as dependent on heroin at the time of death. There was a significant increase in the age of cases over the study period and the proportion of cases that were employed. Fatalities predominantly occurred in home settings (61%). No intervention occurred in 79% of cases. Fifty deaths (5%) occurred in the month following release from prison, 16 of which occurred the first 24 hours after release. Morphine concentrations rose from 0.24 mg/l in 1992 to 0.38 mg/l in 1996. Seventy six percent of cases involved heroin in combination with other drugs: alcohol (46%), benzodiazepines (27%), antidepressants (7%) and cocaine (7%). In only 24% of cases was morphine the sole drug detected. Males were significantly more likely to have alcohol detected at autopsy (49 vs. 24%), while females were more likely to have benzodiazepines detected (41 vs. 17%). The median blood morphine concentration among cases in which alcohol was detected was significantly lower than other cases (0.27 vs. 0.39 mg/l). It is concluded that heroin-related deaths continued to rise throughout the study period, and that deaths were predominantly among older, untreated males. Despite the rise in blood morphine concentrations, polydrug use remained the predominant toxicological pattern.