Developments in the Epidemiology of Drug-Related Mortality in the United Kingdom over the last 30 Years: Continuity, Change, Consistency, and Challenge

Abstract

This thesis has its roots in and has organically grown out of the author’s previous professional civil service roles and academic and related activities associated with drug indicators since 1987. However, the focus of the present work is on providing a description and critical evaluation of developments in the epidemiology of drug-related mortality in the United Kingdom (UK) over the past three decades or so, and how an understanding of that knowledge has been disseminated and impacted policy and practice. First recognised as a major issue by the Advisory Council on the Misuse of Drugs in its seminal 2000 report, a lot of effort and resources have gone into tackling drug-related mortality. Despite this, the number of deaths rose steeply across the UK up until very recently; however, until the effects of the Covid-19 pandemic on death registrations are known and understood, the apparent levelling off may be an artefact of reporting. The increased numbers of deaths occurred against a background of improvements in case identification, investigation, and reporting - with associated public health and law enforcement initiatives. As far as the author can establish, there has been no overview or in-depth study of drug-related deaths (DRDs) in the UK and associated issues. Furthermore, there has been no detailed description of the processes underlying how such events have been identified, investigated, recorded, analysed and the information about them disseminated. Moreover, no study of a comprehensive nature has been conducted into what has been occurring, what insights are now available that were non-existent three decades ago, and what uses have been made of that knowledge. The aim of the programme of research reported here was to rectify the lack of knowledge in these areas and to start to fill the deficits and provide a framework for future research. Relevant chapters contain details of the materials and methods used rather being presented in a discrete methodology chapter. These choices are based on the author’s knowledge, experiences, expertise with regard to sources, and the availability of methods. These factors have been influenced by what the author regards as appropriate, feasible and relevant. Given the breadth of the data being investigated, generated, analysed and evaluated, a range of methodologies have been utilised. These draw on a range of disciplines, including: criminology, epidemiology, netnography, pathology, pharmacoepidemiology, pharmacovigilance, sociology, surveillance,toxicology, and toxicovigilance. The range of methods used include both quantitative and qualitative ones, i.e., a ‘mixed methods’ approach. Given the specific methods employed, ethics approval was deemed unnecessary by the relevant University of Hertfordshire Ethics Committee with Delegated Authority. The main text of the thesis comprises five distinct parts with individual chapters looking at specific aspects, as outlined below. Each chapter includes information on the materials and methods used, as well as relevant references for ease of use. Part 1 provides details of the structure, with Chapter 1 outlining the context of the study in terms of the varied reasons for drug use, the extent and impact of such use, especially the ultimate outcome - death. The research problem and rationale are then presented (as briefly stated above). This is followed by: an explanation of key terms; delimitation of the study’s scope, aims and objectives; philosophical and paradigmatic perspectives (epistemology, ontology, axiology, etc.). A brief discussion about materials and methods is presented. Part 2 focuses on the processes from identification and investigation of DRDs through to dissemination of statistics and other information on these events. Chapter 2 explains the wide variety of terms which are used to describe deaths associated with use of drugs, both direct and indirect. International approaches to defining deaths associated with drug use are outlined and are illustrated using descriptions of different DRD categories. Definitions and classifications of such deaths in the UK context are presented; their uses are summarised. Using a literature review approach, Chapter 3 describes the range of sources relating to information about DRDs and other drug indicators in the UK, together with methods of data generation, curation, and analyses in general terms. Appendix A provides information on the National Programme on Substance Abuse Deaths (NPSAD). A narrative description is provided, using a literature approach, in Chapter 4 of changes in key areas associated with the investigation and recording of deaths: scene of incident preservation and recording by the police; identification of potential cases on behalf of a Coroner/Medical Examiner or Procurator Fiscal; toxicological investigation; autopsy/postmortem examination; understanding mechanisms of death; recording of substances on Medical Certificates of Cause of Death; International Classification of Diseases (ICD) coding; and counting cases. Central to such processes is toxicology in drug screening, identification, recording, and reporting of deaths. Without accurate and complete information on toxicology being passed to the UK’s General Mortality Registers (GMRs), our understanding of the number and nature of drug-related deaths will remain partial and incomplete. Opportunities to address the current identified deficits are examined. Employing personal knowledge, information from key informants and relevant literature, Chapter 5 presents a chronological narrative description of developments in what has been reported and published in concerning DRDs, both at the sub-national, national (UK) and international levels. Data collection, analyses, sharing, and dissemination are linked together by the theme of collaboration. The number and range of agencies involved in these activities have increased across the UK. Some exercises were ‘one-off’ endeavours, whereas others (e.g., Special Mortality Registers (SMRs) such as NPSAD) have evolved and established themselves as ‘go-to’ places to obtain timely, accurate and detailed information. These resources and their major outputs are described here, together with emerging improvements and recent challenges (e.g., Covid-19). Part 3 comprises two chapters which present empirically-based detailed epidemiological descriptions, analyses and evaluations of DRDs in the UK. Using a range of in-depth secondary analyses of UK GMR data, Chapter 6 explores the evolution of UK DRDs at a macro level, both at sub-national and UK levels. The key elements investigated include: demographic and other characteristics of those dying of drug-related poisoning; spatial, temporal and geographical characteristics of deaths; changes over time; (toxicological) nature of the drugs involved in deaths; and characteristics of deaths associated with drug poisoning. Where possible, comparisons are made between the constituent parts of the UK; similarities as well as differences are noted and explored. Data published by the UK GMRs are limited in their ability to provide an in-depth low- or micro- level understanding of all the factors at work in DRDs and the interplay between them. To do this, it is necessary to drill down into the more comprehensive and detailed information available to the SMRs (such as NPSAD and the Scottish Drug-Related Deaths Database). Such an approach facilitates researchers in identifying those groups ‘at risk’ of a DRD, especially through focusing on key aspects: demographic; social; economic, health (physical and mental); and substances. A series of analyses are presented which demonstrate what sorts of insight can be gained. These analyses are a combination of ad hoc literature reviews on specific factors, secondary analyses of existing published GMR data, and reference to relevant studies, using NPSAD and other resources, undertaken by the author and colleagues in the past and for the purposes of this programme of research. These 15 selected studies cover a range of substances and scenarios with which the author was directly involved. His on-going data collection activities and planned outputs are also outlined. Chapter 7 illustrates how low- or micro- level information can give granularity to some of the phenomena described in GMR publications as well as an understanding of other factors to which such sources do not have access. Higher- or macro- level factors influencing DRDs at national (and international) level are considered in Part 4. Such external factors provide the backdrop against which local scenarios are enacted. Part 4 looks at the statistical inter-relationship(s) between DRDs, using data from the UK GMRs, and other relevant key ‘drug indicators’ at a population level for the 1990 to 2022 period. The nine main drug indicators cover the following facets: death numbers; price; purity, drug offenders/offences; prevalence of use; availability; treatment for dependence; hospital admissions; and accesses to poison information. For each of these, the specific measures and data sources (including limitations such as availability) are described. The drug classes and specific index substances are listed (see below). A description of the Excel database designed by the author to hold the data is described. This is followed by an explanation of the statistical approach used to for investigate the relationships between indicators, i.e., the Pearson correlation coefficient with two-tailed tests, as employed by the author’s previous studies. The correlation coefficient matrices can be found in Appendix F. Chapter 9 looks at the commonest drug groups implicated in UK DRDs: opiates/opioids (heroin/morphine and methadone) and hypnotics/sedatives (barbiturates and benzodiazepines – especially diazepam and temazepam). The results are accompanied by short commentaries and comparisons with other studies (where they exist) that used the same approach; this applies to the other groups examined. Heroin and morphine are often taken together with benzodiazepines. Indeed, the latter are often a key feature of opiate/opioid related fatalities, but are also part of a polysubstance phenomenon including stimulants. Some clear narratives are provided most of the substances examined in this chapter. However, information for barbiturates is less clear; but this drug class is no longer of concern compared to the past. Chapter 10 focuses on the main UK stimulants: cocaine, ‘crack’, amphetamines, ‘ecstasy’ (MDMA), and mephedrone/synthetic cathinones. The effects of individual stimulants overlap to some extent but there are differences between them and their mode/route of administration. These specific stimulants are often taken together with opiates/opioids and benzodiazepines. Stimulants are part of the polysubstance nature of UK drug-related deaths. Some clear narratives are provided for some of these substances i.e., cocaine/crack and amphetamine, whilst the others warrant further investigation. The paucity of data points across time and a limited range of indicators restricted what investigations could be conducted concerning mephedrone/synthetic cathinones. The remaining drug classes and index substances are the subject of Chapter 11: cannabinoids (cannabis); anaesthetics/dissociatives (ketamine); Central Nervous System (CNS) depressants (GHB/GBL), and Novel Psychoactive Substances (NPS). Unlike the previous drug classes/index substances, there are no clear-cut narratives to present - even for cannabis. The other drug groups and index drugs considered in this chapter warrant more exploration. Paucity of adequate data points across time and a limited range of indicators restricted what could be said about NPS, and a lack of published figures for England and Wales separately prevented clearer narratives for GHB/GBL and ketamine being presented. Chapter 12 comprises a brief discussion and summary of the main findings of detailed in Chapters 9 to 11. The statistical relationships between DRDs and eight other drug indicators have been explored using the Pearson correlation coefficient across 15 drug classes/index substances. Clear narratives have been provided for some of these substances (e.g., heroin/morphine, methadone, diazepam, temazepam, cocaine/crack, amphetamine) whilst others merit further investigation (e.g., ecstasy, mephedrone, cannabis, ketamine, GHB/GBL, NPS). The limitations of data availability, accuracy, etc. restricted what analyses could be conducted; see also Chapter 14. The strengths of this type of approach have been demonstrated through these investigations. DRD data can be used together with other drug indicators to provide a more comprehensive and holistic understanding of what higher level or macro factors may play a part in causing or contributing to such deaths, in comparison to local scenarios considered in Chapter 7. The material presented in Part 4 could provide the basis for a monograph or series of individual papers on the main drug groups covered here which would assist in informing policy development, legislation, professional practice, service and treatment provision, etc. Part 5 comprises two chapters which pull together the various themes that have been described in the earlier parts. Chapter 13 concentrates on the practical uses of DRD data, especially in terms of decision-making in different contexts. It is argued that information from empirical research described in and conducted for this thesis can be used to feed into public health scenarios, inform policy advice and formulation, inform drug control and regulation, develop education, and provide evidence-based professional guidance and practice. This is demonstrated by reference to real-life examples taken from activities in which the author and colleagues have been engaged during the course of his doctoral research programme. Further details are given in Appendices C, D, E and G. The final chapter (Chapter 14) briefly summarised the findings of the previous ones, draws together the activities described in them, illustrating how the themes examined are complementary and coalescent. It takes a look back at what has happened over the last three decades and what positive developments have occurred or been put in train. Patterns of drug-related deaths in and across the UK over the past three decades are reviewed. An update on the position regarding UK DRDs registered in 2022 is presented. An attempt has been made to see what the future may hold, including: evolving drug scenarios, future information needs; improving the range of information collected and how DRD patterns can be better understood; and ‘horizon scanning’. The conclusions revisit the themes that run through this thesis: continuity, change, consistency, and challenge. The issues discussed in this thesis, especially those covered in the final chapter, have a resonance with the author’s proposed future activities (see Chapters 7 and 13). He aims to take forward and complete the activities undertaken with his recent and present ACMD activities, inform the knowledge-base for the various stakeholders identified earlier in this chapter, and provide a solid basis for dealing with future compounds, contexts, changes, and challenges in a multidisciplinary approach. Drug-related deaths will continue, but better identification of risks and harms will hopefully mitigate such final consequences of substance use.