Abstract

The current paper examines critically the literature on deaths attributed to heroin overdose, and examines the characteristics and circumstances of such deaths. In particular, the dominance of the widely held belief that heroin-related fatalities are a consequence of overdose is challenged. Deaths attributed to overdose represented in the literature are typically older, heroin-dependent males not in drug treatment at the time of death. Fatalities involving only heroin appear to form a minority of overdose occasions, the presence of other drugs (primarily central nervous system depressants such as alcohol and benzodiazepines) being commonly detected at autopsy. Furthermore, deaths attributed to overdose are likely to have morphine levels no higher than those who survive, or heroin users who die from other causes. It is concluded that the term overdose is, in many cases, a misleading term, since it implies the same mechanism of death in all cases, an implication that is neither clinically useful nor consistent with published data. Implications for the prevention of heroin-related deaths are discussed.

Introduction

An enormous amount of literature has been generated in recent years on the links between injecting drug use and HIV infection (cf. Freidman & Des Jarlais, 1991). Substantially less attention has been given, however, to deaths attributed to heroin overdose. This is despite the fact that, in many countries, deaths attributed to overdose remain the most common cause of death among heroin users. This paper examines what is currently known about the characteristics of overdose victims, the circumstances of overdose, the causes of heroin overdose and what implications arise for the reduction of heroin-related mortality.

Mortality and Heroin Use

It is well documented that heroin users are at substantially greater risk of premature mortality than their non-heroin-using peers. Longitudinal studies indicate yearly mortality rates of between 1% and 3% among heroin users (Vaillant, 1973; Joe, Lehman & Simpson, 1982; Bucknall & Robertson, 1986; Joe & Simpson, 1987; Haarstrup & Jepsen, 1988; Tunving, 1988; Segest, Mygind & Bay, 1990; Engstrom et al., 1991; Oppenheimer et al., 1994). The excess mortality rates among heroin users in these studies have been variously estimated to be between six and 20 times those expected among peers of the same age and gender. The causes of this excess mortality are manifold, including HIV/AIDS, the hepatitises and violence. Despite the HIV pandemic among injecting drug users, deaths attributed to overdose remain a major cause of mortality for heroin users, and in many countries is the leading cause of death (Joe et al, 1982; Cottrell, Childs-Clarke & Ghodse, 1985; Bucknall & Robertson, 1986; Joe & Simpson, 1987; Perucci et al., 1991; Eskild et al, 1993; Frischer et al, 1993; Oppenheimer et al, 1994). In Australia, the incidence of opioid related deaths increased by 180% between 1981 and 1990 (Commonwealth Department of Human Services and Health).

The Pharmacology of Heroin

Heroin (diacetylmorphine) is rapidly hydrolysed to 6-monoacetylmorphine which in turn is hydrolysed to morphine following intravenous administration in humans (Goodman & Gilman, 1991). The blood concentration of morphine, the metabolite of heroin, depends on route of administration, drug dose, body weight, time elapsed since the last dose and individual pharmacokinetics (Aderjan et al, 1995). Heroin is mainly excreted in the urine as free and conjugated morphine.

Cardinal signs of heroin toxicity include reduced level of consciousness from drowsiness or a stuporous state to coma, pinpoint pupils and a depressed respiratory rate. Cyanosis, hypotension, bradycardia, hypothermia may also be present. Death is usually due to respiratory failure (Goodman & Gilman, 1991).

Characteristics of Fatalities Attributed to Overdose

Males are typically over-represented in fatalities attributed to overdose (Cherubin et al, 1972; Harlow, 1990; Frischer et al, 1993; National Institute on Drug Abuse, 1994; Zador, Sunjic & Darke, 1996), ranging up to over 80% of recorded fatalities in some studies (Cherubin et al., 1972; Zador et al., 1996). This is not surprising, given the over-representation of males among heroin users throughout the world (e.g. Ball & Ross, 1991 - Griffiths et al, 1994; Darke & Hall, 1995).

Contrary to popular belief, the 'typical' overdose victim is not a young novice or inexperienced user. Rather, the average age of death reported is in the late twenties and early thirties. Examples include: 27 (Swensen, 1988; Frischer et al, 1993), 28 (Cherubin et al., 1972), 30 (Ruttenber & Luke, 1984; Cottrell et al., 1985; Zador et al, 1996), 33 (Harlow, 1990) and 38 (Oppenheimer et al, 1994). Consistent with this have been studies linking mortality with longer heroin-using careers (Davoli et al, 1993; Eskild et al, 1993). Zador et al, (1996) classified cases according to dependence status. Eighty per cent of deaths were classified as dependent, regular users. Given that the mean age of death reported in most studies is approximately 30 years, and that heroin-using careers typically start in the late teens (Kandel & Logan, 1984), most fatal cases have been using heroin for a considerable amount of time prior to death. They do not, on the whole, appear to be novice users but older dependent heroin users.

Deaths do occur among recreational heroin users. An over-representation of weekend fatalities has been reported by some (Manning & Ingraham, 1983; Ruttenber & Luke, 1984; Swensen, 1988), but not all, studies (Zador et al., 1996). Such an over-representation indicates that at least a proportion of deaths are due to recreational users. Zador et al. (1996) classified 17% of fatalities as recreational heroin users. Thus, while such fatalities do occur, they do not appear to comprise the majority of fatal overdose cases.

Circumstances of Fatalities Attributed to Overdose

Most studies to date have focused on the toxicology and epidemiology of fatalities attributed to overdose. A few, however, have examined some of the surrounding circumstances (Garriot & Sturner, 1973; Monforte, 1977; Drew, 1982; Manning et al, 1983; Walsh, 1991; Zador et al, 1996).

There is evidence that the majority of deaths attributed to overdose occur in the company of others (Drew, 1982; Manning et al., 1983; Walsh, 1991; Zador et al, 1996). Others were present at the time of death in 58% of cases reported by Zador et al., (1996). Similar studies have reported the presence of others in 61% (Walsh, 1991), 79% (Drew, 1982) and "more than half" (Manning et al, 1983).

Another variable of interest, the interval of time between the final injection of heroin and death, has been estimated in several studies (Garriot & Sturner, 1973; Monforte, 1977; Nakamura, 1978; Manning et al., 1983; Zador et al., 1996). Instant death following heroin administration does not appear to be the norm. Manning et al. (1983) reported that only 23% of cases collapsed immediately after injection. Only 14% of cases in the study by Zador et al. (1996) were classified as instant, with 22% estimated to have died over a period of time longer than 3 hours. An interval of more than 3 hours was reported in over half (52%) of cases studied by Garriot & Sturner (1973), while Nakamura (1978) reported 44% of cases having an interval of greater than 2 hours. Although figures were not given, Monforte (1977) also commented that instant deaths in the cases he examined appeared rare. The fact that most heroin-related fatalities appear to occur over a period of time presents an important opportunity for intervention.

Witnesses to fatal overdose (commonly other heroin users), however, appear reluctant to seek assistance (e.g. Louria, Hensle & Rose, 1967; Drew, 1983; Manning et al., 1983; Zador et al., 1996). Manning et al. (1983) reported that, in 42% of cases, help was only sought 3 hours after the final injection and that other remedies, such as cold showers and injections of homemade saline, were attempted prior to seeking help. In only 10% of fatal cases in Zador et al. (1995) was medical assistance sought prior to death: there was no intervention before death in 79% of cases.

Mechanisms of Heroin-Caused Deaths

There are several hypotheses about the mechanism of heroin-caused death. The evidence supporting these theories will now be examined.

True overdose

The classical depiction of a fatal 'overdose', as the result of a quantity or quality (purity) of heroin in excess of the person's current tolerance to the drug, is the most long-standing and widely accepted explanation for death due to heroin. If this were the case, one might expect to find relatively high blood levels of morphine at autopsy in people whose tolerance had not diminished. However, morphine levels in studies where they have been reported have been skewed towards the lower end of the range (e.g. Monforte, 1977; Chan et al., 1988; Fugelstad, 1994; Kintz et al., 1995; Zador et al., 1996). Monforte (1977) found 74% of fatal heroin overdose cases had blood levels no higher than those detected in a similar group of heroin users who died of causes other than overdose (e.g. trauma, homicide). Referring to fatal 'overdose' cases in that study, Monforte (1977) commented that:

... one must conclude that in the great majority of cases death was not a result of a toxic quantity of morphine in the blood (p. 720)

A third of cases in Zador et al. (1996) had morphine levels below the toxic level for opioid-naive individuals employed by the analytical laboratories that conducted the toxocological analyses (0. 16 mg/L). Furthermore, many fatal 'overdose' cases are likely to have morphine levels no higher than those of survivors of 'overdose' (Fugelstad, 1994; Gutierrez-Cebollada et al., 1994; Aderjan et al., 1995).

Contaminants

This theory argues that death is not a consequence of any pharmacological activity of heroin per se, but is due to the presence of toxic contaminants in the heroin (e.g. Louria et al., 1967; Ruttenber & Luke, 1984). Toxicological analyses have detected the presence of contaminants, usually quinine, either in heroin samples or at autopsy (Cherubin et al., 1972; Monforte, 1977; Ruttenber &Luke, 1984). Cherubin et al., (1972) reported the presence of quinine in 19% of fatal New York City cases. Ruttenber & Luke (1984) reported a relationship between heroin deaths and the amount of quinine in street packages of heroin. Monforte (1977) reported the presence of quinine in 57% of cases in Michigan. However, in almost all cases, quinine levels at autopsy were well within therapeutic levels.

In general, studies outside the eastern United States do not report the detection of impurities, or of quinine in particular (Nakamura, 1978; Chan et al., 1988; Walsh, 1991; Wahbah, Winek & Rosin, 1993; Fugelstad, 1994; Risser & Schneider, 1994; Zador et al., 1996). Zador et al. (1996) found no evidence of contaminants in injecting equipment or at autopsy of the 152 heroin-related deaths examined. Impurities may play a role in a proportion of heroin-related deaths. It would appear from the international literature, however, that this role may be rela-tively minor, and possibly subject to regional variation.

A number of papers have speculated on the clinical and aetiological significance of pulmonary oedema (e.g. Cherubin et al., 1972; Force, Fisher & Millar, 1973; Byers et al., 1975). However, although congested lungs and histopathological evidence of pulmonary oedema are frequently reported at autopsy in cases of heroin-related deaths, these are non-specific findings commonly documented in many cases of death due to respiratory failure.

A new perspective: polydrug use theory

A theory that warrants consideration in its own right is 'polydrug use theory'. Concomitant use of other drugs (polydrug use), particularly central nervous system (CNS) depressants such as alcohol and benzodiazepines, appears to be a common practice among heroin users. Co-administration of other depressant drugs can substantially increase the likelihood of a fatal outcome following injection of heroin, due to the potentiation of the respiratory depressant effects of heroin. Thus, in the presence of other CNS depressant drugs a 'normal' or usual dose of heroin may prove fatal.

In support of a polydrug mechanism is the frequently documented finding that cases where morphine only has been detected at autopsy appear to represent a minority of heroin fatalities (e.g. Richards, Reed & Cravey, 1976; Monforte, 1977; Nakamura, 1978; Manning & Ingraham, 1983; Chan et al., 1988; Steentoft, Worm & Christenson, 1988; Wahbah et al., 1993; Oppenheimer et al., 1994; Risser & Schneider, 1994; Zador et al., 1996). Monforte (1977) reported morphine as the sole drug detected at autopsy in only 23% of cases. Other studies have reported morphine as the sole drug detected at autopsy in 27% (Zador et al., 1996), 35% (Richards et al., 1976), 38% (Wahbah et al., 1993), 40% (Steentoft et al., 1988) and 48% (Nakamura, 1978) of cases. Considering the polydrug using patterns of heroin users, this is not surprising (e.g. Clayton, 1986; Darke & Hall, 1995).

However, a large number of studies have reported the finding of multiple drug use at the time of death. Proportions of fatalities attributed to overdose that were positive for alcohol at autopsy include: 29% (Risser & Schneider, 1994), 31% (Nakamura, 1978), 34% (Cherubin et al., 1972), 40% (Baden, 1971), 45% (Zador et al., 1996), 47% (Baselt et al., 1975), 48% (Walsh, 1991), 50% (Garriot & Sturner, 1973), 68% (Manning & Ingraham, 1983; Monforte, 1977), 74% (Goldberger et al., 1994; Ruttenber & Luke, 1984) and 75% (Fugelstad, 1994). Studies that have reported mean BACs indicate high levels of alcohol intoxication: 0.09 g/ 100mL (Ruttenber & Luke, 1984), 0.10g/ 100mL (Manning & Ingraham, 1983; Ruttenber, Kalter & Santinga, 1990), 0.14g/ 100 mL (Walsh, 1991; Zador et al., 1996) and 0.16g/100mL (Goldberger et al., 1994). A fifth (22%) of cases in Zador et al. (1996) had BACs above 0.20 g/100 mL.

Furthermore, blood morphine levels have been reported to be substantially lower in cases where alcohol is detected than in cases where it is not, although the clinical significance of this finding remains unclear (Richards et al., 1976; Chan et al., 1988; Steentoft et al., 1988; Ruttenber et al., 1990; Zador et al., 1996). Zador et al., (1996) found that the median blood morphine level of 0.17 mg/L in alcohol positive cases was significantly lower than the level of 0.34 mg/L for morphine only cases. Chan et al., (1988) made similar findings, with mean blood morphine levels of 0.43 mg/L and 0.52 mg/L, respectively, and observed that the mean blood morphine level declined with increased blood alcohol. Morphine levels were also noted to be lower in the presence of other drugs at autopsy. Steentoft et al., (1988) reported the median morphine concentrations in cases where alcohol was detected as 0.4 umol/kg compared to 0.7 umol/kg in cases where alcohol was not present, while Richards et al., (1976) reported median levels of 0.4 mg/L and 0.6 mg/L, respectively. Ruttenber et al., (1990) reported a significant inverse correlation between blood alcohol and blood morphine concentrations (r= -0.14) in a study of 505 heroin-related deaths positive for alcohol at autopsy, as did Zador et al. (1996) (r = - 0.28).

Benzodiazepines have also been frequently noted at autopsy (Richards et al., 1976; Monforte, 1977; Chan et al., 1988; Fugelstad, 1994; Zador et al., 1996). Monforte (1977) reported 12% of fatalities as positive for diazepam at autopsy. Fifty-five per cent of a sample of Swedish fatalities attributed to heroin overdose were positive for benzodiazepines (Fugelstad, 1994), as were 27% of cases in Zador et al. (1996) and 22% of cases reported in Richards et al. (1976).

Is the Term 'Overdose' a Misnomer?

In a substantial proportion of cases, blood morphine levels alone cannot account for the fatal outcome of a heroin 'overdose'. It appears that a great many 'overdoses' are in fact fatalities due to multiple drug use. Furthermore, many cases of apparent heroin overdose have either blood levels at the low end of the range, or at levels no higher than for survivors of 'overdose' or heroin dependent users who die of other causes. In many, perhaps the majority, of cases, it may be that heroin is no more than a contributory cause of death, a point that has been raised by other authors (Monforte, 1977; Manning et al., 1983; Ruttenber et al., 1990; Fugelstad, 1994). For a substantial number of heroin-related fatalities, then, heroin 'overdose' may be a misnomer. To attribute the cause of these deaths to 'heroin overdose' ignores the likely causal contribution of other drugs to the mechanism of death.

The Prevention of Heroin-Related Deaths

The research conducted to date has implications for the reduction of heroin-related mortality. An important finding from the literature is the major contributory role of other CNS depressants in 'heroin' overdoses. The reduction of concomitant use of alcohol and/or benzodiazepines with heroin could be expected to reduce the frequency of heroin-related deaths. Interventions targeted towards reduction of concurrent depressant drug use seem warranted.

Treatment, in particular methadone maintenance, appears to substantially reduce the relative risk of 'overdose' (Joe et al., 1982; Gronbladh, Ohland & Gunne, 1990; Segest et al., 1990; Perucci et al., 1991; Davoli et al., 1993; Caplehorn et al., 1994; Fugelstad et al., 1995). Fugelstad et al., (1995) found a three times relative risk of death attributed to overdose for those who have never been in methadone maintenance, relative to those currently in methadone maintenance. Only 2% of heroin-related deaths in 1992 in New South Wales, Australia, were in methadone maintenance at the time of death, and 75% had never been in methadone treatment (Zador et al., 1996). The recruitment and retention of clients in treatment becomes crucial in light of the risk of overdose outside the treatment setting.

It would seem that a large proportion of deaths attributed to overdose occur in the presence of other heroin users, who are often reluctant to seek help. Interventions to address the fears preventing users from calling an ambulance need to be considered and implemented. Heroin users need to be encouraged to call an ambulance as soon as they recognize the signs of acute narcosis. Police also should be approached to refrain from making arrests at sites of suspected overdoses, as this dissuades heroin users from seeking help. The training of heroin users in resuscitation techniques should also be considered.

One behavioural factor that may become of increasing relevance in relation to overdose is route of administration. In the last decade, the smoking of heroin appears to have become more widespread as the preferred route of administration (e.g. Grund, 1993; Griffiths et al., 1994). Smoking heroin may be a less dangerous route of administration because the drug effect is achieved by repeated small doses rather than a single injection. In a Dutch study of non-fatal overdoses, only 6% of Surinamese heroin users reported having overdosed, compared to 29% of Dutch born users (cf. Grund, 1993). The relative levels of injecting for these groups were 4% and 37%, respectively, suggesting a link with route of administration and overdose.

Research Implications

Several implications for research arise from the literature on deaths attributed to overdose. The exact nature of the relationship between heroin and other CNS depressants in the aetiology of deaths attributed to overdose would appear of primary importance. Specifically, more information on multiple drug use behaviours at the time of death would enhance our knowledge of the risk factors for overdose.

Very little information is available on blood morphine levels in living, intoxicated heroin users. Research into toxicological differences between this group and cases of fatal 'overdose' would substantially contribute to our knowledge and understanding of opioid-related fatalities.

Finally, improved classification and reporting of opioid-related deaths would greatly assist research into the epidemiology and aetiology of deaths attributed to overdose. Deaths involving heroin may be reported using less than clinically useful terms such as 'narcotism' (Cherubin et al., 1972; Zador et al., 1996). Official statistics only record cause of death as due to opiates, and do not report type of opiate or whether multiple drugs were detected. Improved classification and reporting might include documenting the cause of death as, for instance, morphine only, com-bined effects of heroin and alcohol, etc.

Conclusions

The 'typical' heroin overdose death is of an older, heroin-dependent male who is not currently in drug treatment, i.e. an experienced heroin user, rather than a naive or recreational user. The circumstances of such a death are likely to involve the use of other CNS depressants taken in conjunction with heroin. A 'true overdose', involving only heroin, appears to form a minority of deaths attributed to overdose. It is perhaps time to reconsider the term 'overdose' in view of the reported morphine levels in fatalities and the frequently detected presence of other drugs, a point that has been made by other authors (Manning et al., 1983; Fugelstad, 1994). Continuing use of this term to explain all heroin-related deaths is inadequate because it implies that an amount of heroin in excess of the person's tolerance for opioids is the underlying cause of death in all cases. The implication of this term is neither clinically useful, nor scientifically correct, for a substantial proportion of fatalities, as it ignores the contribution of other drugs to the mechanism of death.

For these reasons, continued utilization of the term 'overdose' to cover all heroin-related fatalities may be counterproductive in developing strategies to reduce the morbidity and mortality associated with heroin. The evidence from the literature implies 'overdose' may correctly apply to only a minority of heroin-related deaths. We suggest that, in many cases, alternatives such as 'multiple drug toxicity' may more accurately reflect the cause of death. An understanding of the complexity of the aetiology of heroin-related deaths may enable strategies to reduce such mortality to be better informed, and thus to be more effective.

Acknowledgments

The authors wish to thank Joanne Ross and Wayne Hall for their help in the preparation of this paper.

References

ADERJAN, R., HOEMANN, S., SCHMITT, G. & SKOPP, G. (1995) Morphine and morphine glucuronides in serum of heroin consumers and in heroin-related deaths determined by HPLC with native fluorescence detection, Journal of Analytical Toxicology 19, 163-168.

BADEN, M. M. (1971) Narcotic abuse: a medical examiner's view, Legal Medicine Annual (New York, Appleton Century Crofts).

BALL, J. C. & Ross, A. (1991) The Effectiveness of Methadone Maintenance Treatment (Baltimore, Springer-Verlag).

BASELT, R. C., ALLISON, D. J., WRIGHT, J. A. SCANNELL, J. R. & STEPHENS, B. G. (1975) Acute heroin fatalities in San Francisco, Western Journal of Medicine, 122, 455.

BUCKNALL, A. B. V. & ROBERTSON, J. R. (1986) Deaths of heroin users in a general practice, Journal of the Royal College of General Practitioners, 36, 120-122.

BYERS, J. M., SOIN, J. S., FISHER, R. S. & HUTCHINS, G. M. (1975) Acute pulminary alveolitis in narcotic abuse, Archives of Pathology, 99, 273-277.

CAPLEHORN, J. R. M., DALTON, M. S. Y. M., CLUFF, & PETRENAS, A. M. (1994) Retention in methadone maintenance and heroin addicts' risk of death, Addiction, 89, 203-207.

CHAN, L. T. F., PROLOV, T. & VERMA, S. C. C. (1988) Morphine tissue concentrations in fatal cases in New South Wales, 1986-1987, Paper presented at the 9th Australian and New Zealand Forensic Science Society Symposium, Brisbane, 1988.

CHERUBIN, C., McCUSKER, J., BADEN, M., KAVALER, F. & AMSEL, Z. (1972) The epidemiology of death in narcotic addicts, American Journal of Epidemiology, 96 11-22.

CLAYTON, R. R. (1986) Multiple drug use epidemiology, correlates and consequences, Recent Developmenu in Alcoholism, 4, 7-38.

COMMONWEALTH DEPARTMENT OF HUMAN SERVICES AND HEALTH (1994) Statistical Update, Issue 3, March 1994 (Canberra, Australian Government Printing Service).

COTTRELL, D., CHILDS-CLARKE, A. & GHODSE, A. H. (1985) British opiate addicts: an I 1-year follow-up, British Journal of Psychiatry, 146, 448-450.

DARKE, S. & HALL, W. (1995) Levels and correlates of polydrug use among heroin users and regular amphetamine users, Drug and Alcohol Dependence, 39, 231-235.

DAVOLI, M., PERUCCI, C. A., FORASTIERE, F. et al. (1993) Risk factors for overdose mortality: a case control study within a cohort of intravenous drug users, International Journal of Epidemiology, 22, 273-277.

DREW, L. R. H. (1982) Avoidable deaths from drug intoxication, Medical Journal of Australia, 2, 215.

ENGTSROM, A., ADAMSSON, C. M., Al I BECK, P. & RYDBERG, W. (1991) Mortality in patients with substance abuse: a follow-up in Stockholm County, 1973-1984, International Journal of the Addictions, 26 91-106.

ESKILD, A., MAGNUS, P., SAMUELSON, S. 0., SOHOLBERG, C. & KITTELSEN, P. (1993) Differences in mortality rates and causes of death between HIV positive and HIV negative intravenous drug users, International Journal of Epidemiology, 22, 315-320.

FORCE, E. E., FISHER, R. S. & MILLAR, J. W. (1973) Epidemiological and ecological study of risk factors for narcotic overdose, Archives of Environmental Health, 26, 111-119.

FREIDMAN, S. R. & DES JARLAIS, D. C. (199 1) HIV among drug injectors: the epidemic and the response, AIDS Care, 3, 239-250.

FRISCHER, M., BLOOR, M., GOLDBERG, D., CLARK, J., GREEN, S. & McKEGANY, N. (1993) Mortality among injecting drug users: a critical reappraisal, Journal of Epidemiology and Community Health, 47, 59-63.

FUGELSTAD, A. (1994) Heroin deaths in Stockholm, 1986-199 1, Paper presented at the Fifth International Conference on the Reduction of Drug Related Harm, Toronto, 1994.

FUGELSTAD A., RAJS, J. & DeVERDIER, M. G. (1995) Mortality among HIV-infected intravenous drug addicts in Stockholm in relation to methadone treatment, Addiction, 90, 711-716.

GARRIOT, J. C. & STURNER, W. Q. (1973) Morphine concentrations and survival periods in acute heroin fatalities, New England Journal of Medicine, 289, 1276-1278.

GOLDBERGER, B. A., CONE, E. J., GRANT, T. M., CAPLAN, Y. H., LEVINE, B. S. & SMIALEK, J. E. (1994) Disposition of heroin and its metabolites in heroin-related deaths, Journal of Analytical Toxicology, 18, 22-28.

GOODMAN& GILMAN (199 1) The Pharmacological Basis of Therapeutics, 8th edn (New York and Oxford Pergamon Press)

GRIFFITHS, P., GOSSOP, M., POWIS, B. & STRANG, J.
(1994) Transitions in patterns of heroin administration: a study of heroin chasers and heroin injectors, Addiction, 89, 301-309.

GRONBLADH, L., OHLAND, L. S. & GUNNE, L. M. (1990) Mortality in heroin addiction: impact of methadone treatment, Acta Psychiatrica Scandinavia, 82, 223-227.

GRUND, J.-P. C. (1993) Drug use as a Social Ritual. Functionality, Symbolism and Determinants of Self-reg-ulation (Rotterdam, University of Rotterdam).

GUTIERREZ-CEBOLLADA J., DE LA TORRE, R., ORTUNO, J., GARCES, J. & CAMI, J. (1994) Psychotropic drug consumption and other factors associated with heroin overdose, Drug and Alcohol Dependence, 35, 169-174.

HAARSTRUP, S. & JEPSON, P. W. (1988) Eleven year follow-up of 300 young opioid addicts, Acta Psychiatrica Scandinavia, 77, 22-26.

HARLOW, K. C. (1990) Patterns of rates of mortality from narcotics and cocaine overdose in Texas, 1976-1987, Public Health Reports, 195, 455-462.

JOE, G. W., LEHMAN, W. & SIMPSON, D. D. (1982) Addict death rates during a four-years post-treatment follow-up, American Journal of Public Health, 72, 703-709.

JOE, G. W. & SIMPSON, D. D. (1987) Mortality rates among opioid addicts in a longitudinal study, American Journal of Public Health, 77, 347-348.

KAN DEL, D. B. & LOGAN, J. A. (1984) Patterns of drug use from adolescence to young adulthood: period of risk for initiation, continued use and discontinuation, American Journal of Public Health, 74, 660-666.

KINTZ, P., MANGIN, P., LUGNIER, A. A. & CHAUMONT, A. J. (1989) Toxicological data after heroin overdose, Human Toxicology, 8, 487-489.

LOURIA, D. B., HENSLE, T. & ROSE, J. (1967) The major medical complications of heroin addiction, Annals of Internal Medicine, 67, 1-22.

MANNING, F. J. & INGRAHAM, L. H. (1983) Drug "overdoses" among U.S. soldiers in Europe, 1978-1979. 1. Demographics and toxicology, International Journal of the Addictions, 18, 89-98.

MANNING, F. J., INGRAHAM, L. H., DEROUIN, E. M., VAUGHN, M. S., KUKURA, F. C. & ST MICHEL, G. R. (1983) Drug "overdoses" among U.S. soldiers in Europe, 1978-1979. 11. Psychological autopsies following deaths and near-deaths, International journal of the Addictions, 18, 153-156.

MONFORTE J. R. (1977) Some observations concerning blood morphine concentrations in narcotic addicts, Journal of Forensic Sciences, 22, 718-724.

NAKAMURA, G. R. (1978) Toxicological assessments in acute heroin fatalities, Clinical Toxicology, 13, 75-87.

NATIONAL INSTITUTE ON DRUG ABUSE (1994) Epidemiologic Trends in Drug Abuse, December 1994 (Rockville, US Department of Health and Human Services).

OPPENHEIMER, E., TOBUTT, C., TAYLOR, C. & ANDREW, T. (1994) Death and survival in a cohort of heroin addicts from London clinics: a 22 year follow-up, Addiction, 89, 1299-1308.

PERUCCI, C. A., DAVOLI, M., RAPITI, E., ABENI, D. D. & FORASTIERI, F. (1991) Mortality of intravenous drug users in Rome: a cohort study, American Journal of Public Health, 81, 1307-1310.

RICHARDS, R. G., REED, D. & CRAVEY, R. H. (1976) Death from intravenously administered narcotics: a study of 114 cases, Journal of Forensic Sciences, 21, 467-482.

RISSER, D. & SCHNEIDER, B. (1994) Drug-related deaths between 1985 and 1992 examined at the Institute of Forensic Medicine in Vienna, Austria, Addiction, 89, 851-857.

RUTTENBER, A. J. & LUKE, J. L. (1984) Heroin-related deaths: new epidemiologic insights, Science, 226, 14-20.

RUTTENBER, A. J., KALTER, H. 0. & SANTINGA, P. (1990) The role of ethanol abuse in the etiology of heroin-related death, Journal for Forensic Sciences, 35, 891-900.

SEGEST, E., MYGIND, 0. & BAY, H. (1990) The influence of prolonged stable methadone maintenance treatment on mortality and employment: an eight year follow-up, International Journal of the Addictions, 25, 53-63.

STEENTOFT, A., WORM, K. & CHRISTENSON, H. (1988) Morphine concentrations in autopsy material from fatal cases after intake of morphine and/or heroin, Journal of the Forensic Science Society, 28, 87-94.

SWENSEN, G. (1988) Opioid deaths in Western Australia: 1974-1984, Australian Drug and Alcohol Review, 7, 118-185.

TUNVING, K. (1988) Fatal outcome in drug addiction, Acta Psychiatrica Scandinavia, 77, 551-566.

VAILLANT, G. (1973) A 20 year follow-up of New York narcotic addicts, Archives of General Psychiatry, 29, 237-241.

WAHBAH, W., WINEK, C. L. & ROZIN, L. (1993) Distribution of morphine in body fluids of heroin users, Journal of Analytical Toxicology, 17, 123-124.

WALSH, R. A. (199 1) Opioid drug accidental deaths in the Newcastle area of New South Wales, 1970-1987, Drug and Alcohol Review, 10, 79-83.

ZADOR, D., SUNJIC, S. & DARKE, S. (1996) Heroin-related deaths in New South Wales, 1992: toxicological findings and circumstances, Medical Journal of Australia, 164, 204-207