Incidence, prevalence, harms and intervention effects for problem and injecting drug use: crime, morbidity & mortality

Drug misuse causes severe problems for society, communities, and families and incurs huge economic costs. UK Government estimates suggest somewhere between #9 and #15 billion, in England and Wales alone, during 2000, and perhaps up to #16 billion during 2003/04. Most of this cost related to an estimated 330,000 people who had very severe problems with addiction to drugs like heroin and crack. The most recent economic estimates (2003/04) suggest that these people accounted for 99% of the total cost and that their crime accounted for 90% of the costs they generated. Drug misuse is also the third most common cause of death for people aged 15-34, mainly because of overdose from drugs like heroin. A lot of taxpayers? money is spent in trying to address these problems, but estimates of the number of drug users and associated problems have not been routinely or consistently available to help governments to make the best policy choices and it is difficult to be sure how certain the estimates are, because they relate to different time-periods. Our research will combine, and then compare, different sorts of information about serious drug problems to get a more exact idea of how many people are involved, how many get involved in crime, what proportion die, and how helpful treatment has been in reducing death and crime. Instead of just relying on how well people can remember their crimes, which is the approach that research studies have taken in the past, we will use treatment records and criminal records to see exactly when they were arrested and if any crimes happened while they being treated, or not. We will use a similar approach to look at death registrations. We will use statistical analysis to improve estimates of how many people have had severe drug problems, to learn whether more of these have died, and from what causes, and to discover whether this problem is worse in some areas than others. Using advanced statistical techniques, we will examine all of the available sources of information about drug misuse and bring these together to test whether they are consistent, to find out why inconsistencies occur, and to adjust the estimates to allow for any inconsistencies that we find. Our findings will provide the Government with much better estimates, to help to make decisions about how to deal with the drug problem and how to spend public money wisely.

Decisions about drug policy are based on estimates of harms, prevalence, and intervention effects that are imprecise, sometimes inconsistent, and lack cross-validation. There is a need to develop more precise, better validated, and consistent estimates of key harms associated with Problem and Injecting Drug Use (P/IDU), and of the effects of treatment intervention, in order to provide a sound basis for policy decisions. The proposed research will examine the prevalence of P/IDU and key harms within three related strands. Strand A will analyse data for a virtual cohort (n=320,000) of PDUs in England, already assembled within work for the Home Office, to explore the timing of crime and mortality events in relation to evolving treatment status. This will provide a novel and more precise assessment of the magnitude and temporal persistence of treatment effects in reducing crime and mortality than has, hitherto, been available. Strand A recognises the cyclical nature of addiction treatment and avoids the difficulties inherent in self-report data. Findings about patients and circumstances that constitute a higher risk in terms of mortality or recidivism will have a direct bearing on treatment policy and practice and will inform work in Strand C. Strand B will apply Bayesian approaches to demographical IDU prevalence estimation with a view to comparison of IDUs? drug-related death rates; and will test the application of this approach to a common set of national sources for Scotland, reflecting the economically efficient approach taken in England Strand B will also explore regional trends in IDUs? drug-related death rates, with England as comparator, and so determine whether IDUs? death-rates by age-group and sex are regionally and nationally consistent. If not, questions about the differential effectiveness of IDU-related harm reduction policies may be posed. Strand C is the first comprehensive attempt simultaneously to model P/IDU prevalence, mortality and crime data together, using a Multi Parameter Evidence Synthesis framework. It will test the consistency of the available data and produce a coherent model, for England, of drug use, mortality and crime, cross-validated by several entirely independent data sources, to better inform policy development. Methods and results will have wider application in areas such as alcohol dependency. We will place particular emphasis on dissemination of results to policy and practitioner audiences, as well as via peer-review journals.