On this page:

National Investigation into Drug Related Deaths in Scotland, 2003

« Previous | Contents | Next »

Listen

Chapter 7: Conclusions and Implications

What is in this chapter? This chapter reviews the findings from individual chapters and draws together a final set of conclusions along with associated implications for service commissioners and providers.

Where did the information come from? The information has been drawn from the "key points" sections from each chapter in the body of the report.

Introduction

This investigation has comprehensively examined the circumstances of death, social circumstances and service contacts of those dying of drug-related causes in Scotland in 2003. It has also considered long-term trends in drug-related deaths in Scotland and has compared the circumstances of the 2003 Scottish deaths with data from a concurrent study into drug-related deaths in London for the same year. Finally the investigation has explored the opinions, beliefs and knowledge of living injectors regarding overdose risk.

The developing evidence-base implies that many drug-related deaths may be preventable and this investigation aimed to identify areas which, if effectively addressed would impact on the rate of drug-related deaths in Scotland.

Each chapter of the report has therefore identified "Key Points" based on the data collected. These are summarised below and have been used to generate implications for service commissioners and providers arising from the findings of this investigation.

Describing the study population and trends over time

Most Scottish drug related deaths in 2003 were male, and in their early 30s (mean age of 32.7 years). Most were considered to have been accidental drug overdoses (based on ICD-10 criteria for drug related deaths), although a sizeable 13% were classified as suicides. These characteristics are consistent with previous published studies into drug related deaths.

Trends in drug-related deaths in Scotland from 1996 to 2003 were complex and showed a great deal of heterogeneity over time in relation to geographical distribution and the involvement of heroin/morphine or methadone. Drug-related deaths involving heroin/morphine had increased at a significantly higher rate than those involving methadone in Scotland during 1996-2003 (13.8% vs -0.4% per year, respectively). In Glasgow, however, deaths involving methadone increased at a higher rate (average 9.7% per year) than deaths involving heroin (which increased at an average 6.5% per year).

Drug-related deaths involving heroin/morphine had increased at a significantly higher rate out-with the main urban centres of Glasgow and Lothian during 1996-2003, and likely relates to the growth of drug use in these areas. Overall, there were twice as many deaths involving heroin/morphine as methadone. However, the ratio of heroin/morphine : methadone overdose deaths varied considerably between geographical areas - for example, from 117 : 26 in Lanarkshire to 21 : 70 in Tayside and 59 : 175 in Lothian.

A higher proportion of methadone compared to heroin/morphine-related deaths in Scotland occurred at the weekend (defined as Friday to Sunday). Drug-related deaths increased at a significantly higher rate among those aged 35-54 compared to 15-24 years during 1996-2003.

Implications

The study and analysis showed important differences in drug-related deaths over time and geographically. However other studies will be required to test potential explanations and hypotheses for these differences. One key question is to what extent do the differences follow, and are due to, the underlying trends and pattern in drug use in Scotland? Or are there differences in the availability and delivery of treatment, and investigation of drug-related deaths, that may have a bearing on the mortality statistics and mortality rate within Scotland.

Toxicological findings and circumstances of death

1. There were missed opportunities to intervene and save the lives of many of the people who died. Less than half had tried CPR and most cases had died by the time the ambulance arrived.

Implications

There is a need to develop and deliver training and education for drug users and their families to increase awareness of the risks of overdose, how to avoid it, how to identify it and how to respond effectively.

2i). Benzodiazepines were the most common drugs detected in drug-related deaths. Few cases were positive for psychostimulants, in particular cocaine. The predominant drugs found at the time of death continue to be heroin/morphine, benzodiazepines and alcohol in Scotland.

2ii). Nearly half of methadone related deaths, and two-thirds of diazepam and dihydrocodeine positive deaths involving illicitly obtained medications.

Implications

The widespread ingestion of benzodiazepines in particular diazepam among drug users (as indicated by the toxicology data for the current sample of deaths, and by the data on trends in drug-using characteristics of cases of drug related death during 1996-2003 in Scotland) is a matter of major concern regardless of whether these drugs were implicated in the cause of these deaths. Benzodiazepines are potential drugs of dependence with risks of both acute and chronic adverse psychological, physical and social sequelae for misusers. More research is needed into the prevention and management of benzodiazepine abuse.

The findings also suggest that illicit manufacture and/or diversion of prescribed drugs is a substantial source of drugs for users, and remains a significant issue for health service providers and others in the field. Where diversion of prescribed drugs e.g. dihydrocodeine or methadone and/or benzodiazepines, may be occurring, there is a need to address prescribing practices in relation to these drugs. However any further analysis of illicit sources of medications would have involved collection of data on the illicit manufacture of prescription medications, which did not form part of this investigation.

3. A relatively high proportion of cases did not inject any drugs prior to overdose and death. A high proportion was also positive on toxicological analysis for only orally consumed drugs. Drug users who do not inject heroin (or other drugs) are also at risk of fatal drug overdose.

Implications

The findings reported here suggest that overdose prevention initiatives should continue to reinforce the message that most fatal drug 'overdoses' are polydrug deaths and alert drug users that for a substantial number of cases, 'lethal' drugs are not necessarily injected drugs.

Services must be aware of the risks of non-injecting, in particular oral drug use, and must continue to warn drug users and the broader community of the dangers of combinations of drug use including alcohol. Drug treatment services may also need to be aware that non-injecting dependent drug (in particular licit drug) and alcohol users, may not be attracted into seeking help because they may not identify with these agencies' illicit drug using and/or injecting clientele.

4. Analysis of cause of death information showed that the same drug-related cause of death may be recorded on the death certificate by forensic authorities in various ways. It also found that while 'opiate intoxication' was the most frequently recorded cause of death for drug related deaths in Scotland, being recorded for 131 (44%) cases, in only 12 (9%) cases was toxicology positive only for opiates. In the vast majority of opiate intoxication cases, toxicology identified other drugs in addition to opiates, most frequently benzodiazepines and alcohol.

Implications

That post-mortem toxicological results are not always reflected in the certified cause of death could suggest either under-reporting on the death certificate of other drugs detected or more likely that other drugs detected were not considered to be implicated in the cause of death.

The task of attributing cause of death to one or more of a number of drugs detected at post-mortem is a complex and inherently fraught one in the case of many drug-related deaths. However, under current cause of death nomenclature, the prevalence of polydrug use, and in particular the widespread involvement of 'licit' orally consumed drugs in cases of drug related death, may be underestimated.

The systematic collation and availability of forensic toxicological data would assist the surveillance and/or analysis of trends in drug-related fatalities. Developing a more standardised, uniform nomenclature for recording drug-related deaths on medical certificates would also improve the monitoring and researching of drug-related deaths.

5. There were a higher number of deaths positive for dihydrocodeine and benzodiazepines in Scotland, and a higher number of deaths positive for cocaine in London. There were a higher number of cases in receipt of dihydrocodeine prescriptions in Scotland, and lower average blood concentrations of methadone and morphine in Scotland.

Implications

These key differences, which might reflect differences in the pattern of drug use and treatment provision between Scotland and London, could be explored further.

6. In Scotland, a higher proportion of deaths (17% vs. 10%) were associated with recent prison release ( i.e. within three months of release) compared with London.

Implications

Part of the difference may possibly be explained by differences in services provision between the two sites. The larger proportion of deaths with a recent prison history in Scotland, which is a recognised risk factor for drug related mortality, is worthy of further investigation.

Social circumstances prior to death

1. Information, as it is currently collected across Scotland is sparse, inconsistent and difficult to cross-reference.

Implications

Use of a standardised, well-validated method of collecting agreed data on all drug deaths would substantially facilitate the identification of relevant social risk factors.

2. The lack of up to date relevant information in many of the casefiles, which would be required to organise an integrated care plan is a concern. The availability of rich, up to date information would allow identification and prioritisation of potential risk factors in this vulnerable population which could reduce future morbidity and mortality and must form part of good practice in the management of drug misusers.

Implications

Staff in all settings should be trained to comprehensively and holistically assess drug misusers and to ensure that regular updates of essential information ( e.g. regarding childcare responsibilities, life events etc.) are recorded. Nationally, standards could be set within the DAT Corporate Action Plan requiring recording of adequate information. Locally, DATs and their health and Local Authority partners could ensure that services are commissioned with clear quality standards and monitoring procedures in place.

Contact with services

1. Most people who died of drug-related causes were known to services and many were accessing more than one. These services were often generic ( i.e. not specialists in the field of substance misuse) and were often accessed in an ad hoc, chaotic manner. Following this contact, most were discharged with inadequate follow up in place or failed to attend any such follow up appointments resulting in discharge and no further action.

Implications

There is a need to increase awareness of this problem and to deliver training and improved coordination of activity for those generic staff most likely to come into contact with this group - General Practitioners, Psychiatric services, Accident & Emergency and Social Work.

2. Medical interventions were only accessed by a minority. When accessed, the intervention was often prescribed methadone. Quality of methadone prescribing was often outside that contained in national practice guidelines. Few of those prescribed methadone were in receipt of any counselling.

Implications

Methadone replacement should be prescribed only in line with the current evidence base - following a full assessment of drug problems and dependency; in adequate doses to meet need; dispensed safely and effectively under supervision until the person is demonstrably stable. Methadone replacement prescribing should be delivered alongside supportive counselling.

3. Many had been in prison in the last 6 months. Ten died within 3 days of release of which 6 were released on a Friday. Transitional Care is not being made available to all who require it. Prisons show varying success regarding take-up rates.

Implications

There is a need to ensure all opportunities are taken to intervene when drug misusers are in prison. In particular, there is a need to ensure that effective communication takes place to ensure imprisonment does not interrupt treatment (if in treatment before incarceration) or that it gives an opportunity to increase access to treatment if not already in contact with services. DAATS should review the care pathways across the community/prison interface and ensure adequate accessibility to services for all newly released prisoners. Prisons should look proactively at how they engage with prisoners at the start of their treatment cycle.

Experiences of overdose survivors

The majority of overdoses occurred amongst those who had taken more than one substance on the day of overdose. Heroin had been taken in all but one overdose. Heroin and diazepam taken together or on the same day was the most common combination, taken in 22 cases. In fifteen overdoses heroin and diazepam were the only drugs consumed; in the remaining five episodes they were taken in combination with one or two other substances and/or alcohol.

Injecting drug users are aware, to some extent, of the overdose risks, although this knowledge is not extensive. Less than half mentioned tolerance as a factor and less than half considered that a mixture of drugs could be risky. There were some misconceptions about overdose risk factors. Some injecting drug users believed that overdose would not occur if heroin was smoked or if it was injected by itself. Injecting with someone else present was the most frequently cited prevention strategy.

Inflicting physical pain was the most common intervention used by injecting drug users. Although putting the overdose victim into the recovery position was cited by half the sample, the infliction of pain was still regarded as the most effective strategy. Half of the sample was worried about having another overdose. Among those who were not worried, some claimed not to care whether they lived or died.

Implications

There is a need to develop and deliver training and education for drug users and their families to increase awareness of the risks of overdose, how to avoid it, how to identify it and how to respond.

Conclusion

This investigation has identified a number of implications for services in the prevention of future drug-related deaths in Scotland. It has highlighted some limitations of fiscal and national registry office data on drug-related mortality and has raised aspects of drug-related deaths requiring further research.

Finally, the report provides some baseline indicators by which the potential effectiveness of interventions to prevent or reduce drug related deaths in future might be measured. Some of these outcome measures might include improved after care from prison, improved delivery of methadone treatment, and increased engagement of drug users with services especially drug treatment services.

« Previous | Contents | Next »

Page updated: Wednesday, August 3, 2005